Codes/2D_VNET.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "accelerator": "GPU",
    "colab": {
      "name": "2D-VNET.ipynb",
      "provenance": [],
      "collapsed_sections": [],
      "machine_shape": "hm"
    },
    "kernelspec": {
      "display_name": "Python 3",
      "name": "python3"
    }
  },
  "cells": [
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "yuhGP_fSBIq0",
        "outputId": "3f58c6c1-b749-472e-a570-b67532dad377"
      },
      "source": [
        "from google.colab import drive\n",
        "import os\n",
        "drive.mount('/content/drive',force_remount=True)\n",
        "os.chdir('/content/drive/My Drive/')"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Mounted at /content/drive\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "T6wEVlw2Gghe",
        "outputId": "eab0bd48-c60f-40dd-8345-525dfc6b7c57"
      },
      "source": [
        "!pip install SimpleITK"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Collecting SimpleITK\n",
            "\u001b[?25l  Downloading https://files.pythonhosted.org/packages/9c/6b/85df5eb3a8059b23a53a9f224476e75473f9bcc0a8583ed1a9c34619f372/SimpleITK-2.0.2-cp37-cp37m-manylinux2010_x86_64.whl (47.4MB)\n",
            "\u001b[K     |████████████████████████████████| 47.4MB 62kB/s \n",
            "\u001b[?25hInstalling collected packages: SimpleITK\n",
            "Successfully installed SimpleITK-2.0.2\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "z4rxB4f2C_eI"
      },
      "source": [
        "import SimpleITK as sitk \n",
        "from tqdm import tqdm\n",
        "import numpy as np\n",
        "\n",
        "from keras.utils.np_utils import to_categorical,normalize\n",
        "from keras.models import Model\n",
        "from keras.layers import MaxPooling2D,add,Input,Dropout,Conv2D,Dense,BatchNormalization,Concatenate,Input,Dropout,Maximum,Activation,Dense,Flatten,UpSampling2D,Conv2DTranspose,Add,Multiply,Lambda,PReLU,concatenate\n",
        "from keras.optimizers import SGD,Adam,RMSprop,Nadam\n",
        "import keras.callbacks as callbacks\n",
        "import keras.initializers as initializers\n",
        "from keras.callbacks import Callback, ModelCheckpoint, LearningRateScheduler\n",
        "from keras import regularizers\n",
        "from keras import backend as K\n",
        "\n",
        "from sklearn.model_selection import train_test_split\n",
        "import matplotlib.pyplot as plt\n",
        "\n",
        "import skimage.io as io\n",
        "import skimage.transform as trans"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "vVLTjIlkOQ24"
      },
      "source": [
        "#Loading data"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "pcrl0XP1F9fQ"
      },
      "source": [
        "path = './HGG1/'"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "u0Cp9hx3GPMr"
      },
      "source": [
        "def load_data(path):\n",
        "  my_dir = sorted(os.listdir(path))\n",
        "  data = []\n",
        "  gt = []\n",
        "  for p in tqdm(my_dir):\n",
        "    data_list = sorted(os.listdir(path+p))\n",
        "    img_itk = sitk.ReadImage(path + p + '/'+ data_list[0])\n",
        "    flair = sitk.GetArrayFromImage(img_itk)\n",
        "    img_itk = sitk.ReadImage(path + p + '/'+ data_list[1])\n",
        "    seg =  sitk.GetArrayFromImage(img_itk)\n",
        "    img_itk = sitk.ReadImage(path + p + '/'+ data_list[2])    \n",
        "    t1 =  sitk.GetArrayFromImage(img_itk)\n",
        "    img_itk = sitk.ReadImage(path + p + '/'+ data_list[3])\n",
        "    t1ce =  sitk.GetArrayFromImage(img_itk)\n",
        "    img_itk = sitk.ReadImage(path + p + '/'+ data_list[4])\n",
        "    t2 =  sitk.GetArrayFromImage(img_itk)\n",
        "    data.append([flair,t1,t1ce,t2])\n",
        "    gt.append(seg)\n",
        "  data = np.asarray(data,dtype=np.float32)\n",
        "  gt = np.asarray(gt,dtype=np.uint8)\n",
        "  return data,gt"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "CyhZuIKwONOt",
        "outputId": "656fb1c6-78f9-4c6f-efe0-bab76d78817a"
      },
      "source": [
        "#HGG\n",
        "data,gt = load_data(path) "
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "100%|██████████| 10/10 [00:44<00:00,  4.48s/it]\n"
          ],
          "name": "stderr"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "ofnRuBS0Y4Da",
        "outputId": "86dff879-54c0-480a-8c23-3cb28b68c16b"
      },
      "source": [
        "data.shape,gt.shape,data.dtype,gt.dtype"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "((10, 4, 155, 240, 240), (10, 155, 240, 240), dtype('float32'), dtype('uint8'))"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 7
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 521
        },
        "id": "zQpbQvSZY--A",
        "outputId": "e4f4287a-13ca-4b26-af2b-cc919c35ba61"
      },
      "source": [
        "\n",
        "plt.imshow(data[0,2,65,:,:])\n",
        "plt.show()\n",
        "plt.imshow(gt[0,65,:,:])\n",
        "plt.show()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "roMBLbffKgsB"
      },
      "source": [
        "data = np.transpose(data,(0,2,3,4,1))"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "DiRnDDVvkYuk",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "60d8b99e-fd71-407e-c53b-52153cc69ce2"
      },
      "source": [
        "data.shape,gt.shape,data.dtype,gt.dtype"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "((10, 155, 240, 240, 4), (10, 155, 240, 240), dtype('float32'), dtype('uint8'))"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 10
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "hDY8ISjFk8jV",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 521
        },
        "outputId": "dbdfe7a7-7cd2-4d55-b63e-ae1ef66f1180"
      },
      "source": [
        "import matplotlib.pyplot as plt\n",
        "plt.imshow(data[0,70,:,:,2])\n",
        "plt.show()\n",
        "plt.imshow(gt[0,70,:,:])\n",
        "plt.show()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "SLNUTlTBsSPY"
      },
      "source": [
        "# Pre-Processing"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "J98P5jKcQlvh",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "6eb7f454-f519-4ad5-af93-06b4636775a7"
      },
      "source": [
        "data.shape,gt.shape"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "((10, 155, 240, 240, 4), (10, 155, 240, 240))"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 12
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "NJSrnmls1sol"
      },
      "source": [
        "data = data[:,30:120,30:222,30:222,:].reshape([-1,192,192,4])   #taking 30-120 slices\n",
        "gt = gt[:,30:120,30:222,30:222].reshape([-1,192,192,1])"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "yG6JZMTK2HRX",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "758906ba-7c0f-4df9-eda0-d2aabd89f61c"
      },
      "source": [
        "data.shape,gt.shape"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "((900, 192, 192, 4), (900, 192, 192, 1))"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 14
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "cEJnl3cCAloP"
      },
      "source": [
        "gt[np.where(gt==4)]=3   #converting ground truth value of 4 to 3 => one hot encoding"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "6R48jIx1N0nv"
      },
      "source": [
        "X_train, X_test, Y_train, Y_test = train_test_split(data, gt, test_size=0.2, random_state=42)\n",
        "X_train, X_val, Y_train, Y_val = train_test_split(X_train, Y_train, test_size=0.25,random_state=42)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "1yN7fCRoTaAQ"
      },
      "source": [
        "Y_train = to_categorical(Y_train)\n",
        "Y_val = to_categorical(Y_val)\n",
        "X_train = (X_train-np.mean(X_train))/np.max(X_train)\n",
        "X_test = (X_test-np.mean(X_test))/np.max(X_test)\n",
        "X_val = (X_val-np.mean(X_val))/np.max(X_val)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "LIYVglRp2B7K",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "236069ee-6560-42f5-9609-59eb5b2624c6"
      },
      "source": [
        "X_train.shape,Y_train.shape, X_val.shape, Y_val.shape, X_test.shape, Y_test.shape"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "((540, 192, 192, 4),\n",
              " (540, 192, 192, 4),\n",
              " (180, 192, 192, 4),\n",
              " (180, 192, 192, 4),\n",
              " (180, 192, 192, 4),\n",
              " (180, 192, 192, 1))"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 18
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "j6z2pMvRD51W",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 521
        },
        "outputId": "749d2da3-32e4-4d8f-def9-6902a5bcab4a"
      },
      "source": [
        "plt.imshow(X_train[0,:,:,1])\n",
        "plt.show()\n",
        "plt.imshow(Y_train[0,:,:,1])\n",
        "plt.show()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAQEAAAD8CAYAAAB3lxGOAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjIsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+WH4yJAAAgAElEQVR4nOy9a6xtWXbX9xvzsdbae59z761bVd3Vbre7aXe3wQ64bRq/gETEgRAUBUgCwZEAJVGMlFgEiSQQEiGUfEGJAUWKhGSUSP5AyBMSklgJxOFD+BDiRyDE725o0273s7qq7nnsvdaac458GHPOvU+5qrrdt25Vd989pKt77nntffc5c8wx/uP//w9RVc5xjnM8veHe7idwjnOc4+2NcxI4xzme8jgngXOc4ymPcxI4xzme8jgngXOc4ymPcxI4xzme8nhiSUBEfqeI/JyIfFRE/uSTepxznOMcjxfyJHgCIuKBnwd+O/BLwI8B36eqP/2mP9g5znGOx4onVQl8B/BRVf0HqroA/xXwu5/QY53jHOd4jAhP6Pu+G/jEyb9/CfjO1/vkQUad2D2hp3KOc5wD4IqXPq+qz7/6/U8qCXzREJHvB74fYGLLd8r3vl1P5RzneCrif9f/7hdf6/1Pqh34JPCek39/fX1fD1X9IVX9iKp+JDI+oadxjnOc44vFk0oCPwZ8UER+jYgMwB8A/voTeqxznOMcjxFPpB1Q1SQiPwD8b4AH/gtV/akn8VjnOMc5Hi+eGCagqj8C/MiT+v7nOMc53pw4MwbPcY6nPM5J4BzneMrjnATOcY6nPM5J4BzneMrjnATOcY6nPN42xuA53t745J/4HpYHSh6Vr/vbyvav/p23+ymd422KcxJ4CsJ/84d48SPPkkcoHlyG8pse8exujxPlk/4dPHjHdwMgBUoU1EOJkLaQtgoCkkCyIAXcAm6Fb/grHyd98pff3v/gOR4rzkngazTyb/t2ircD+9lvHrn67j273YHNsDKvgfvDihflZh64fO8ryPsz3imHNbAdVoIrPJj2/Lp7n+bXbz+BF+XnDy/w0rqlqOPjNw9Zs+fFT30Dl7/0TuInXyZ/9B++3f/tc3wZcU4CX0shgv/G94EIH/1XlTAkSnbALTFmhpAJruDHpX+Jc4WSPaU4nBRKcdzMAyLKfg0UFW7yyMvLhr//2Xexrh7nlP0XNvhdovzTC5/Jjgc//i6+rijkDMtK+tSn37aX4Ry/ujgnga+FcB63mXD3LvmF/+gezhcGSeQs+JCJMTPFxBgS0WdycXhXUBU2MZGDvS9lxxASIspuWDmkwO068NMvvcDnrnbcfm4HoeDGDFkoLw/I/YU4Jm5+S+KnP/Is7pXI8JLjfT94BaqU/R7OC26+ouOcBL4GIv8T38ov/GEhjIlps6AqDCHhnTL4jBOlqJBVcMWxjQuDz6TiKCoAqApL8eRiA6NDCqTsOEhARFmXgNs71AsaFKaMJsc4Ji63B1SFR25iSY7DNvOzf/6b8deeb/rBj5+rgq/wOCeBr9L42H/y3eRnVxBFBMQXnLdy3rnSE8DFMONE2adIToG1ONbiuYgzFyGT1BLB4DKjT7w0bznkgBPlpXVDcCtL9qQ54Gchb0Bnx/jMAREYYmI3LAwuE3zmUcik5BFRZjfxM//+e0Hfy9f/qLL5H/7vt/tlO8drxDkJfJVEeNcL/Mx/8F5IAl7h3owoaBG0CIggAtFnpmHFiRJqyV/ADulQSMURXWbyiY1fSWo3f5DC6BI3fqAgZHUMwaqIXARdHH6BvAFZHd4XntntCa5wOcwMLuFdAWBJniUFZq/olJGgfOJ3euR7vxONijs4PvDH/q+38dU8x2mck8BXcFz9S9/Fyx9yqIO8UfwzewBElJwdWuzgI4p4xblCDJnoClkFEbVDXltyJ9qTw8av7MLM1Toxl4Bz2j8HYBNW1sHjXcG7ALFQIuAV9YrW9uIirGzD0tuKKSSiK0wxkbNjPkScU2RMiFNizKyr5x/9me9BnaIC3/jDnz5PFt7GOCeBr7B49H3fRZpsTv/oA7A+v+CGjI+ZGAohZNbVfmzqBFVBC4hThpCZQsKJsmTPCgRXkHr4vRRiP+yFokJBSOpw6thJYXCJl/MGsKpi9Ik5BVws5EkNDxiswmj4QSqOwWUu4szgMgVhTgF/qdyMkTV7cnbWPlRw8upDt2gRvC/84u97F+NLL/Dw52b83/rJt+V1f5rjnATe5vDPPANf9w40etLFwOe/Tci7gnoFZyQdBFA78AJ4X/DeSu9ShFIP03ZcGENizR7AgEAVoisEKQw+M7jUHzupZy2+3+JJPUsJHFJEakUw2rdCnFKCgqglpfr4uWIKTpSdXwhi7x+cTSO2ceGQIodk48boM2v2DEOqiUE5fMue28VzeH7i3enbcP/n//PkX/hz9Dgngbcp/HPPIjGyvu+dfOEf27JeCIeH9aBlEBU02kHULKR8PKwxJsaQEbGyXESJPnMRF2sBxPXbvz+es1t+G44cgbX4Dgym4nm0Trx02HKzDASfSdmjKqRSW48sKJaEos8AdcJgFYFDGX0iSGGjwlUaGVziIs6kk8d6NE/EkAkhM8+RsjpYhf17Fz7+z0588MUPQcrkn//YW/TTeLrjnATehnDbLVe/9QMcnnHkAdZLYbmvrPeKSbqUfvtTFK06L3VKwuN9QSSxjSvRZ7wUgitMfiUVT66HstQqoI0Iiwqjs4nAUgJr8f05HXIgqWefIoc1kA8DqkKovAJNDr8IZRTikBijVRRttJjUEREGl7gXDgCsagli41ecFFLxzCUYYAnczgM5ecgCRUCF9OzKz/zRB4SXPR/8s58nv/zKW/iTeTrjrCJ8q0OEl/7538DNOx3rhTA/I8wPlfWZgm4zOmV0spJastgBwcpxBJwv5OzIxd5fapvgRG0C4DJzCizJ48Vu5l1YuIg2KrxJA7lVFK7d5p6bdeRmHSgqBFdIyaqJ6ApDSIRpJd3LMJY+ApxC4mKYeTjesvEr+xz5zP4en5kvuR/2PIh7ostEl/GiFAQnlqx8TUwiatOOWOxPS4LAR//EN+N2530UTzrOlcBbHJ/9N74bKUraiolzdkq6KOg2ETcrIRQONwOaBFlt9EdUxBfEae/F1+y5XSPR2zSgjecmb8Bb9Mdb3g6e3dx24zuCFHbBevhDDjxaxs4kLCoMQ6YU4WaJRF9seuCOeEDKvlcYY8UZUnEUhNs08InDMzwbb9j4laLCy+uWOQeCywQxXYJ3pYOLilGYnVPWIZCLEF5xfPRP/wY+9J99gvSJX3oLf0pPV3zZlYCIvEdE/paI/LSI/JSI/Fv1/X9GRD4pIn+3/vldb97T/eoNiQOf/Te/BwTSRlgvYbmnpE299mZPycbLn3YL7mJFLzKyzbghI44ODgIsS+DmMLBmzxRW7g0HdrXfD1K4HGcuhhkR5TYNPFonllr+L5UMtM+R2xRRFXzlA9zMAylbMnjm4haAl683zDcD1JGkc0qpmOVaPC/OO15ZJ5YSWLLnNg28tGz5B7fPMZdAlMzGL4w+UdRxnUZScTw73fD85TXb7cwwJLyvxCenqChlVNTBR7//Pchv/Ja3+kf21MTjVAIJ+OOq+pMicgn8hIj8zfqxv6CqP/j4T+9rI/xzz/Lpf/FDrBcgCjlC2ihlqBOAArII+SZwmwUXCuLAT6nfjiW7DgKuSzASUMnoaEnBoczF46QwhsRyclNnddysnlQcD4Y9SwlcryPBZZZivwLRZ7IaACkCY0wGGGZPSh4Wh6xCycIYV3bDCsCcAnM6/hq1iiQXxy7OvLJuWHzozxFscuDw9TkULqeZ/RpYUmBdPSU5CJYE/I2NS//R77rPC89+hPg3fvwt+7k9LfFlJwFV/RTwqfr2lYj8DLaD8BwnEd7/Pj7zT76L/TvFDj0V7xtORoCtDy6CHjw5OFzM+KD4SgJqZJ4QcicKeX9E/40HoGQ1cC4VdwcQzMWxiueQ7Ue+TxGwUWCbJGj9XFVYk2dfYscbsgr+1pFiYLkMRF9Ykr9TzoPxEoaQWVybBjjcaG3B6FMlLBW+MO+Y63NxYhyHUhyLBgMJ/cl4VGC9UD7zkYHnxu9g+p/O9OM3M94UTEBE3gd8G/B3gN8M/ICI/CHgx7Fq4aXX+Jo7uwi/FsN/8P18/rveyaMPgIqSJ5uzSxG0HmrEUH8q848isEIRhziQorSuzbW+f1w7qNZu3qKCQ8lI1wLYFxmCX8Tev5RAEBvxzTlQiuEDoU4Yos9ksVFezs4YiENijRF1HsnC/nakFMG5ylxU4yoA5GCPK3LkDwwVjwguU2rrAUZlXoonn7Q5PYpAsYSpUdEA+xcKnw2Br7v5jYT/4yee0E/t6YvHng6IyAXw3wN/TFUfAX8R+Ebgw1il8Ode6+u+1ncRhvd8PV/4znfw0rdAiUreFHQo6KCUsdgv9slNTih2+xXqARBKsoOYs1RSkP24NsPKZliZYmLw5hFQkD6SC64QahIItT0YfSJWUM5VZqDj+Pgiap8XE8GbGCiEzBjtfX7KlE1BQ6EksRYBOlB5qhbW2kbMq7UKhxyZS2CfI/scWYtncIldnBlrgmgVi0itkOr302BtU5nstZufy/zyPz7ivvXXPaGf3NMXj5UERCRiCeAvq+pfBVDVz6hqVtUC/CXgOx7/aX51hb93j8//tvfw0jcJJSolKjqcnBJv1FuiAWBSxH4SLSmc8ASaPsD7YtqAStKJ3ijC27gweevPl2LI/+RNQNSAwEYSasmhVJ4AWJJo3gJzCtVXwBN9YawCIu8U5zM6FPxFYns5s9vMDCExxcQ0rIxjYqj8ARElF8e8BrIKQXKlKbtOGJp8qtwG4zu0/7Zziq8JUaNSttleO6cQLHmuO+Vj3/eA8K4X3oKf5td+PM50QID/HPgZVf3zJ+9/18mn/V7g//vyn95XX0gIvPh7voXrd9f5fq4HPJTjDVdvekQtEYy5i4DkIsEmI74QppXNdmEc1woKwhByH615Z1RgoLL+XGfwDS7zYNgz+NTbgI4RIH2cBwbiJT3WBa4e4pQda+UkmAgo43zmYpp5ZrtnU41K7m8O3N8cmGoC8M5wBu9L5QO4nqCKVq1CxSGW4vGV0hyqAOpidyDulsodMA6BjLnjBOqNWfmz/+77cJeXb/0P+WssHgcT+M3AHwT+voj83fq+PwV8n4h8GPt1/zjwRx7rGX6VxUv/8m/i8IxQzxe91V2dJYHBymeyMeQa8CVeifU2BTuIwedO221cflVhOyw4lE1Yj0KgSvJpB2twmct4YJ8jByLBLb0Hp8BeI7H26I1hGF0xhD77rkjcL5GcHSW7rgLcL7F7FTQ9QlHhZh25XgcE2ETDLS6GmSmspGKJYPIr9+MBJ4WrdSLXZOREGWPqAKdRlR1p9vb6VAxFnaKjWsLK8HN/9pv5tf/hPyR/5rNvyc/3azEeZzrwt+m/6nfiqV5CmiMGxjnQYD0tBVxylKkgvpj+v1UEggFgxaMxGw3Y6Z1bfsmeWZQYMkOlCT8YTVbc3IGaTsChd34qG7+yqe2CfcHYQcQ5hztfm8oRgGzPoSWfooLm6j3gM0v2RO+5Nxz6ob6IM7vV8J1WdZh5yVJ5CqHrH0aX2LvM5XDo7891iuF95v7GqMd7BtMW1ETJUCyxZsGtDgp87I9+I9/4lx+Qf/rn3+wf51MRZ8bgmxUiXP++7yRthRJBfU0EDnBQorUDWgSWo8YfBcmua/Rb5OJYoOr5C1N1B74YZuvvqwnILIHbFFlKMIWgWCLYhRmg9/6DS0QpPIh71mDl+CFHqyCkkNSxZs9c6cYtAU0xsYpSirAkx7oEgi9sYmITVpYcuJKRjV+JUtgGMyppjwdGKHJSuBcPRJeZi+cmXRBc5n488OK8IxdHPql4dnHhZh7QWjGJM0ykCOTFIYtND9wqSIZf/qee44VNRH/ip96SH/fXUpy1A29CuGli/d5vZ9lVv34PeaqMt1isGmjc+Hb4u0S4/luhZGFeg7HxGtGnOAQYfWYMqeoDDOHvjy9awTftpfVNGtnnaB/H+vKC4QFtTDi4dOdrojdD0lAR/+5MVPEIESUtxg1Q6Lf6IUVeWSau0tidikaXeRBv2YXZ2IK19bhJg3EHatm/z5GkR3zD1+eRtGoXxoSLpVcr2iooFdwiuBWkCOsOPv0991l/x0ee6M/6azHOlcBjhru8pPy693H7zogGKMGWdaStomMFA8FuM1HjBMR6gJNpA1Qq4JXtpnVOu1FoNwRpakCkj/bWCrSFyve38Z8SXa4+AY4oqZfmLSk0ko59XSapTRGiy+yGhUM6gojt8UUUcVCScDhExriyjUdWInhcOU5AmqZgF2aKCnMFBg85du+BlpSClC5mAhh9Mk/E4tiMKwepvgnZ1IySBVFbouL3BrzmCdZLZb04E4p+tXFOAo8R7vIS3v/1PPrAjhLEyn8PZcQSQAMB6+xfV4fEggzZvAHFQcaqATXfgJwd6+oryp76zQgc0X9M/9+fB3ZQ28y9SYSdlDtJI6v0iuD06wL1AHoDHlexry/6KyEf8UpJNkZMxeHw/fudPp9rHZmqhVmUTKyEpKLCkj2XQXsymmoLYW2DMvmVueoboHERHDljwqrSXtO6MclDHo1PkO4XPvms433X3352KfoS49wOfJnhdjtLAB+6x1pxgJMKvQJ/9e1W+iej5IpXXFBczDb6iuXIEQBKdp2JF6orEND75iX7rgZ8dWQVDhXwW04IOgDxVTdu4ThV2FavQKkehLke1lwMsCt1pCnO/jSm4hFLMGJQm2QUNTXhTRpxotwL+y5wWkpgPklirYJpY06wqiSGzLya5XljS7ZxSxu9lgFru4LxCjQW9J0zH/s+j//g+7/Mn+7TFeck8GVG/vAHeenX32e+58hD9e4XSwThRoivePwrAbn13RMArycEICuvRRQXC36b8JtEiNn++NIXgbSSeRPMRGQpoSr2Qu/nG0PwIiw8M+yNp4/W0ZyZebTvYxiBHdLWCrh6U2/Dwr3xYKPC6hq8rqFbm5XFU1bXE1KjIAOsNUG1SuSQI58+XPIoTYwu8SDeMlVp8fV6nFKUqnfYhJXBWXsy+dQ9E6aYTDuhlVvhsL+hkrEMFwCQoeBrcv2ZP/6cSbHP8YZxbge+zFgvI4eHDvUcy9LBRn9utRtKkCoSdNYeeEWCgYNaf6HFYyBXvWW9NzPR7biYc5A7koFaedzK9CJ2iE5v9+YZuHa2YG0l1FGyQPcViOxT7C1EKlbWDz5DhnvTgTV7XtlPla0Ia21bELpfoHeFWHGIxjlYsydU85DJJ6Lk3jI8M+y5jHOvTkafKlho48TJr31qcVnXpc2pKhlbAqAeeq3Tl1ZxOZBQ8F7JTtHTyuwcrxvnJPBlxM2/8J3cvOD7GBAH2Unf+psnu51KrCVqOFEMAuKOYiConUNxeJc7y85VwozjWHqvxdedAWt3Emqefg1822s0xZ5LzOX4403FcZsHXjzsTGdQPQvX7FmzZxfn7gU4+Mz9Yc8hR7IKV4eReY5ocfgpo8WAujV7NmHttOVDjja289pbFSeF6zxynUcDKl1mIws7v7Cq45V1gxPl/mC8gJs08GiZcHUc6l3hCzdbVgwzaePVvCuGqWBcDB0U3eQ+2RBXW4ZzfNE4J4FfZVz//u9ivjwyAqWAZFvjXTx2U7lapo4FWgLgyHqjynaVOt6Ld01Dx5h6IoDjzb8LS+fgnwDxOL/2doAS7iDvN2ngeh1Zi2f0iaSOz1/vUGw6cDEutYc3Rp9DOaTIUo7If/QZRjgcIhTHuF0JvrDURFIa4FjxhZt16KNMMCzidKSZ1DPnwFoTRSM0ZRX2EjtHweTO5m0wr6EuWTFFoSQzYpW1GrJerAxTqspGoazOku85vmick8CvMkowJmCJ9rYUbMYfDKAqA5RQFYJCnWnTgb9W8jtnYFebv4sogrHxrJx3lTrctghJXw/Wn0vFA9qN37T6LVpLMPm1H8g1e8aY+r7CTbCPJXVcr2PnHAyOyus3rT9ACp6cYRpWLseFqYKJQzUvLQj75lTkDHuIUljV4euUwrCHjPPKmsY+udjXqqNUwVFLSNFncnDchox4cxoi06cDZVPQKeOjvaZGNpJeef3Sv/fdfMN/+vcoNzdP7pfiqzzOSeBXEfrd30qOQh6k3/wuV817g1j12K9Krl4BTmrfqvXw10Mv1ha0orWJgwD8iVcANL8A4bTNbWh6mx70Gxm54yTcxoANBJxCYi2OweeO5MMRP2hkosmv7H0krXawQzDE3on18veGA0M1Eg2tIvB3RUmnZifrCR8hV18B71MfXS7Fs+Qjw7H93UaJItjhbroLZ6NYv01d0gymRBRnnIzbb0hIPP+av1GcX50vJUQIv+a9vPj+DXmqYymPAWQem69XgE+y3VZSmj5I0IqmabEWwLT3guqxBbDSG6jqwCYGAu7M4VtV4OtYrf05/Zx2o9qta4YiAK7YzdpK+MHlo+bg5LEaYDj50jcYtd0G3mlnF7aRX9MrNDPT1ka0v1tCagpGp6Vbno+V6lz05OO1OmjJrSUS54rxMarzkApINCA1hIJWwLUlAYKe16J/CXFOAl9CSIjcfNPz5FG6NoB6GTUOgOkE7op3AJscaHUSSkJZPEnFgMEAIsVMNVWgOJIKYzzOzPshVyHhOgjokFq2m1a/3azN+bdgOwZiJQwBrGJA3pp9dxHqVQS2P2AtBhS2nYU2fSioF9bGXfAVnOSYfIKYXXjDBaDe4s3D0GUcav4FJ1iHEzVHJDm+fdrSLCUQnU1LGocgCYb8C7hQ+jKUlty0vr4acxc9neP145wEvliIINNov+AB0mS3vpQjOUj9yee3MRq1LTCym4FUWdDZkZNQgqNoRhVCMFygeYmc9v2nkYsDR6cRQyP8uOMhqx7/QQobv1DUGUeg0ngfDPs7ngKpinuCKEGsCmh7CdqhaxqGUhwF2ETtI8XTx3R1JtesxU0SbI9VsrALCxvfNiBFLuPhiBlUpD+pY6hVTqqJcPKJy8EEUY/qVycMX5k2i21PdoVQKwozS/X4UEjlPCH4YnFOAl8k3MUF6cMfII+OtBE0YLd7rtyAQO//iceR4alQSN2Jr2AVvyiFIo62GbBhBc33H47JIBXXFYan5Xuoq8Vaz70vkVBKF+1EySC5e/s1TCDUCuFRGjuXv/XgRR0vzxtLKqKsVMbgSV66WSJONl0F2KqNVP0CwcaFjQmIHkHL3ipgiSSe9PINhwDDJ1Kpa9DUEoEbrS24CamCj+a5ALYOvT1HJ/ZaFtrSlnM18EZxTgJfJESEtPVGDR7qre+gnKwLK1G7qCVHjpMB6tueLhKiAEMxF6E6Mmwmnc5ZNZCy40DoJT6YpwDQ147Z+44WYQA7v7Cp48ILb8KdjMOple1RzADkJo3si2c8Fe3UtwefuDcemFMghIQ/Kd+zmrdgcz4W2XUcoNRkEaWwYPRhnP3fWrUBNi4cnW0rejDs7f3qmHPg0TqRiuPheNuT3ZyDJcZqlBp95sEmd6HUnI2uHH3p7QqY3mDej+jiOLOG3jjOSeANwj//POs3vZvlfmC5EPJkQpV287t8pAuXSakXrxlgFDGPgMBxw3AbE4aCa0pCtcrAeytpG1AIrfwvfVIgUMvv497BdtimE+OQrVt6AoBj9RDFkPylBK7T0I1LmzvRWm3JWizZ9hqOwRD8/TxwuBkIQ+5kp2Zh3mJfbPlIKo6SjizHyZtuwQxRPZsqEspOuvVYq24aR6ElJDjiB0Hs43MKJH+0Oy+1MgBbolLWgM4e1tojnON145wEvlh4QQXyxrwCtPGEHBRnkwAV4wigtVLw9sto40G6k7A4JQy531bmrNvm2/QEUIqjlMYfcDihTwzg2AqcSojHOqobnbEFb8uAp1Rk/cg4vM0DQY79enC5VwEtKYAtFZGT1qPtJdDZm+9gdkzRevXWlkQpzHW+38hJcNxcnE9amst46C1KUzM2fOH0wL/aOen0/383YYpNYhWWFFiWAFmQ9dwKfLE4J4E3CAmeNHnSJOTh7seaa1CjBuNaG9BQQbpWoJX9rrUAd9SGp84iddhw8gvfNQYNH1CHZM/kbcSXqNJidWxk6YliclYZRMlHFN9ZddH0/OEEhe9rz2syeWXe1PVkhVzZgD1m0yFcjZOx+zaY03A79MFm/g3LaM+vrTBv7UFLDKcJoH2sVTgtmuuQd6VTk8GSVXtubRlK32IkekedeY7XjsdOAiLyceAK43ElVf2IiDwE/mvgfZjZ6O9/rQUkX8nhdjv03gUluj4SLCc0VBVO5Kv18A924LWCf+Kt97e5td34qncXbcgJa/DUXPTV0ZJBf5ujV/9plHpwPZiY52R0YQy+TBHrrZ0UbtJ4h77rpHDhZ+YpVEmyu1vyi3HytbYzuSL47UAHl9mw1kRgX5eKqy5Gp0nHVb6Daa5DZR028LBFcKXTh32tGHDFxq71/5+q/iFnm2BoEdMZtOnMOd4w3qxa6bep6odVtXk7/UngR1X1g8CP1n9/VYW7vCA93JHHYxtQYv2gNefVRKT2+1EJYzYrrFDMK8C1BZuFEFoSOD7GHdZgveWamxAc24PG6mslcpsgNAvvdpMCdw59Yw4eii38aCSj0aXOJ1jVxoRezJfwws+MLvGO6Yr7w4FtWJh8YqyYBd7aHDdmhldpHMCAv403OvG9aF+/DUu3DoOj2Ul/nidVQiNJNWyiSY3LyaE/JUgpVgGk+ifnmhi10rVfI6Ge4248qYbpdwM/XN/+YeD3PKHHeSIhIcBmIm8DaSOUIOYW5KotYD38nSrste4ONMcgcbX0r+2ByN0JgPPFKoBOIa7dQ20bThMB3OUFgI3PzKHnuOjTid3yo0vdQhysvJ/c2nkEQMcIWow+sfMLW7cwutQTyWU48GDYm8Q3rGZ2OmYIio+2naiZnsQ6dhzdShDDGe7FA4NLvUXp/IP6/XNlE+5z7ItSTg9480s4JTSd4gNzsiWmpsI0+7GczO+gvjC9YjnH68ebkQQU+Bsi8hN1vyDAO+vCUoBPA+98Ex7nLQv/zneQn7tH2nhKOE4FRK0lKFENAGwHtZgdd1rDkShUaTil2YYAACAASURBVK7el+oRYD7+zeOi8QJcddEN1UjUu0KqZa21CMdWoZF8wgmCf8oohCMWsKpVAFCnApLZ+pn7fs/WLf1rH8Q99+OejV9Y1aTKz8VrLrxVBW17kDEMC86bE9JusxwXn1TS0OhWMnarN7fji2Bjy1M+A1gCOK1ITiPUaqJNAaLP3SQlq2OfYic1FeXIQKxtQLd05ywn/lLizQAGf4uqflJE3gH8TRH52dMPqqrK6bVW4yt6IekQSReR9cKRB2sDtJJ+1FWb6wIgHRwso5C9s6pgzNXwgy4bbvsE7W/XJcTmIJQ7Y9BoueU473YN4Te+/iasd/0FK1V4c6IU9HX2PrmVtXiy2Nu+HtTmJbB1Sx8jTm7lwh+YxHr3KJnM8ea+TQNTSOw2C0s86vansHai0aqeKOZFcJsHxmDbka/T0ElJWYWLsBjpp+IRjSLdMISHoyn+tmHheh3Zp3in5Wm6h7YROauQk+9YjI1z6uE/Mwa/aDx2ElDVT9a/Pysifw3bPfgZEXmXqn6qriX7FethVPWHgB8CuCcPv6IgXN2MlFgps0FIG8CBnwU3Sy8AtBKFbG7oevtZFgdDocRSJa5tBGfiFl/Ve+3gm41WVfK1BaNqcmJorULp1mB92Wh9X1tG6ilc+ANRMtd54jqZYccoC6Nbe0VgB/yI+LcJwlwi9+Oee27Pi1wwl8hYF5cGV9hGcyJuy0uzCnMODD5zVQ/0g7g3bkHdSlRE6q0euv9AlOPC1FSO69Kb5Pl0a3E4aYNUqwlKcf1zoi+k7PE1kZbizEtABKKJinDnRPBG8bgLSXcictneBn4HtnvwrwN/uH7aHwb+x8d5nLcy/PPPky9G1NXb3hkHwB/AH+RICDo1Ea1JwR8El8zwgtXssUsW0up7JSACIeTuKeDFPAM2g5XxrbRfqognupOtwhXRN5KQ79ZiTZTjRFlLON7I4Vj6zyVyUDtsWzezdVb+r+otIZwAdS1RrHr0JjxdajoEO5wXw3LX36BLhwu+4gQXfuZB3JvRSbUcb2DkqQ+CrTMzrOOqbjFqDkjasQTHWnkDvr4ugzdsIsaMDwUfMm4wj0E3ZJiOrdM5XjsetxJ4J/DXbDcpAfgvVfV/FZEfA/4bEfnXgF8Efv9jPs5bFjKNlMFbmR8gbe2297euk4GaerB4wwYEcIscwUIHBAMGtbr1umBTgJwFVU8IhvK7umEIjv19+0Uf5Lhv4Nj7u+Mtqg5yAJ+IVDMStXFZlNxHhO2mB/AUBmeHtMl5ixc8hdEn5hK5komdm1m97yBjEHvMTVy5Xow00STEr/Y1uO/33JbBHlcKq3qCy2yr9NiLWiJQ16uMV087uvy4JgKwx2mpas1GRhqDtSKqQva5G6Dmqh7MWfjYv/PNfOAvfZL08X/0RH5nvtrjsZKAqv4D4Ftf4/0vAt/7ON/77QqdBtIusNzzzPeENNF1AdRRfWs7cceE0EaIOpQjRfgEOGxf1JyFoj/21Y0efAqdNBPPweUu+W3uPX1PQA3HrxwRruptKiCmNWgHsr09Ouvl1+JJtRoAyAg3ZWTnZjzFcIWTsePtMDCnwFocIlatWAJw9e2qcZDjc4yS2fi1Khp9Nxl9dTTpcp8CYLsUmhfinZ8T1TuAI7/COyEXJZfCKlZBeA/6ax9RLr7CcKevoDgzBl8VuhlIO89yIawXQomKW+lioT52FqsCtE4K1FXSUCxI9RXsnoJo9+wPITN0h567UEjrc5ucONQqoSvwTkr2VhmM3lqF4KzPX4vcPYDO3H59X0DicFgiuPB20DO21GR17cZ1fbIwupVDicwEomQu48yLbgfFXIU3cTVGYfMsOKk+DiX2JDS6xOJMI9BaHSfHSuI0TseAQ9232NqkE3Fmf8z++RVnyaU5ER1t287m+q8f5yRwEm6aKFO0seBgqkHqNKAngVAPvYc8KGWqZqLQKwB3IrABqi7g7ljQ3n+XaNPWffXyn6OU2LnjvsEGCLZdAUEyHqParuqJPrP1tr7c1z+tLD+91S/9gUt/4LYMHErs3gQFI/RMsjLJyrVMZJwpEtu2o7hwuw54KazFG+gXjAF4KJH78ZZclYuN8APcKf+X6jtwunikcQgu6oixnHAemmNSY0926/WTNgHorMxm3ZaLg3LGBl4vzkngJOQ9X0fe2g2onqN0eKGrAbuhaF02SlTEl2OJ4Ozm8d7KflXI2XVqcCn1F/QEzGsrucvJL7aZatjqbV+NQAMQXOoz91MXnn6zFiMIbd3C1s29zAeYZO1VQpRmFWYThdsycJvHfnDbDd4+99If2DoD9JrVeBMVNT1CUkeQ3PGA+35/55Zvk43Jr1xX8C8Vf4JxWDJpSsLLONumpbZ+rR5u7yzh5TUy+uPhbq9jX6fenIiSQ/RX8hHOYXFOAqcRA7kKhtYLSFOl7/rqIhxrSb0pMDVqMHdovs5nvD8y/uzwN/KQvX8ImcHn3k/3bT4+221eFXpFheKEAQghGZcfc+gZXGKfI06PZKGbPDK4xKHErhPYurmLgABrGdT3G7b1/XbrJ54Pjzho5Cpv6mKR1MeOV3nC1U1Ik18p0W7m0dumpOZE5Cm8krc8DNccNPYKZK304PZnKf6uzyHHfQu3OuBEWUow09ITNWGub8eKo7QNSK4d+uxwrrCuoXIz3Nlr8A3inAROomwieRRyFQyB/Z0nyFM5moNEUwf6cBQM+VCIMfUdgs39tpWlpThUFe/sl/hqtZc+q/SvGWOqIpzjwG7Jdjs25FxVWKPnueG6gm2VgVdLdU/pB7tXCdi4blVD+3duZlFT301uxVVcIEpiUc9NGYmSmNzaS/CMEF3iws9sgo0zm7hnE9YTIo/jws84KSwaOGjsQN/pWBCqorEuUnFo344EVIt1Iwm1qmPOphKMPhMw2vCaPd4VpmAj1LViFcbJKOS6Yfm8juz145wETiJfDJQo5E3FAyryn8fqDlR7fomlS1lLruu9RNE6h/e+mKNOY7FV/cC6RtxmIdddhG2UlZJnGJJhBcWxGVYjEZ30vLdrJLpCcaVTbdt6r1byB5d5lDY8HG54V3ypjuI8BVtWcpU3ZGzr8KqeS3fA1YSxw8Z3vUoQe/zJrXhRJl15OW+ZS+DesO9chegycwo2s3dmWfYgDjwM1wySuGVgLZ6bNPLSsukHe65l/uUJHboZiSw59LagkaGaLNlVoC+rsCRLAH0XYuUUBFdIuCooMr4G5awheL04JwEAEfwHfg37ZyJpEjorV048AgpQ369qjlUF152rmuVWm/+3EVY7oCI2GWgMuKOxyBErAAi+9L62/cLHRgd2hV1Yeg+9qr9j+Amw8QuewlXZ8Ky/ZsVufxsNJtYyEt1S+/7EPX8gSjppF4TnwyNyXTruUbyseEoXGEUxf8RtWHA5cCDecRL2YpXFTRl5JW15OW25SiNX68T1MnQMYQqJpYROO24RXGY6+Xd0meRdZVe6bnjS/AVzceTeftUJQXbGyci2Ev7sLvT6cU4CNdYX7nN4UE95w9IK5jacoG8TcopmoXgHVTYs1UsvRnMPTrW8bx74p4sxigo5++P4SrQnjnmNxEopHmoiyXUsNoVkisCwdkou0D36zLFX2Tj7vLlEbmTsj9tu/IPG3i54mrrPdRBwJ4bKH8pQ2wbPqgEnhUHSEYDkyPRL6rsPwKbanBV1HDTUFsQmAbdr5Powomp+BsEV5hRwwSqAbjnulANmN96wgOYkZHgAOMxOrL1v4YjN+DqK7VVAljMm8AZxTgIA4jg8N7BeCH6v3ezHZSCD1q06ir3dloxqsY1ALtp0oOn/T41D2kF/9Zz7zsPXjye1Qx850odPY6ikoSi1oigQXS2ny9C3/IyS+hSgHe6C8QMmWavpyDEBRGeMQzAhUVHH5Nv8/2jo0b5fdJl9rroCdRyItF2EQKcqr9XqvG8XSoF19f3/bM9L7uwuOJ12wJFD0CqdpqwEq6aW5CtYaMKrNmYVsFbsLCD6onFOAs4T3vN1zPcdeQS3Vt+ADG6uPXk0y3Co48HTjUCYKEYmJa3mdd9MQuBY7ntXcE0ZWBdltJZBFbIcf4lPN/icjhDbvN2JAYClLuxoPP+1DCwl4KVw6fc4jgkA4KCVyntit+OlEDmOAwcyWQpbyXwhX9ypEuDIBIx1I7I7sQprz3WsoGTDK06TWTk5lKf+h8fno3csxzqVuoqP7DW1Q5+yr5MA4x7kIsbirK9tUTHm5llO/Ibx1CcBf7Hjxd/6buYHVaQy2W4Bl6AtGpVTd61kNvaqmJFlFsMHnN2tOZuIJcbjmnGoJapPvVo4LLFbYrUedghGJW4sQTgSYUT0zq3pTnAEKmW3GYdEyVy6A6t6lm7g4ZhLrCNDrci/TQS6EakGDigZwaNsnbUGGbGWwFAQwJh8Q1UYNtvx1grMJXadwFxCdwoaggl9cnbHLUW+qQldV0k6URZqq1OTbAMOoVUAob923XGII9bStxANmSIK/kwZfL146pMA3jPfc6yX4OejQMhVpaA4rNR3dwv5tpJciqDJkR0QCzhBi6OUtn3YzDhO2W0pmQmnEYeoI0W7AaUYT1587hVAi+bVZwrBW+4HK7lvy8BMIEjmXjCL7qsyMcnKIJlcD+5YdQDPumsu3Z6X846rvGF1ga3MREnc6MBNGVndgUnWngAABslEl3gYb7jO451qJFVdwMaXLlxqI8tU24HSxqDeEeum42YacpuHii1oH32esgVHb8mq1BYguMK6enIKDONa3YWqRbsz7oAOQgmFFBy4cxJ4vTgnATAX3kFRqfJhgdwWjlamoPq7bYFksyJXr51E1NaQ51RddouYUWkxrYAXW611szfATvXIamuW48U5IJGyEWQE21bcyttYkffbPNbxXekMv30ZeJQmtm7hpowsJyWMKQwDo7NR31JHge4EM9jJTK66gUlWHpWJ2zLWl6iwauh9/tYtRJd5ad12g9LRGbjXREfmgnTs8Q3cLH30OdbV6EvxTD4xBfs++xw78Wg4UUw2zgAYduJc4epmYl3rQhQxP4E1H6unRtzSM0/gdeOpTgL++ee5/Y73mWdgAFL9G0w2XL0E6j87Nbi3sQKlioiI5QhCta25age9sdqyCjeHwUr77Eirh2KuxMVbNaBq8+6UPGm1KUKZVjZx7SacV2kiOWctR00Cz8VrDt6IOQ/DNZNbucobwCqArZsprNyUkUESRV13D7opIwciURKDZN4RrnB1zBirLgEsUUSXII9c+LkvEWnqwaSeQZIRi5qzscqdQ91K+kMKlNDWj+cqkKofy7WF8GvfvOS466y8pIAT2G0W5kq8ukNRlqNT0/n8v3E81UlAgidtHGlLVwqWoXoCCMdNwyq2UBRsJW77WDQBkUzZRofJSDbij8Bgzq7fyKdW4qU48yYsxitu+veSnRmWtoTjjh4DJrBxREwxeJ3Hfqi8FO57W+uVcSwaailvXP/iji5CWzfzwN3yuXyPz6VLomQe+Ns6PjQs4KocJ/WLmhtx0wOs1WikiZGajflaAvscuc1jL/cv/MzGr7yM7TGYwnFT0ubkbaBrJVr1EKRQnDEH00lbVPS4dMS5wmZcKI1aXIytOUZLMNeHkWXxZ+3AG8RTnQTwnjQaOaij/638b+vD1Hp/DVpdhsrRTcgBg60Ua4j16VZisIOfkmnvRZRljkY+EuMbsDo0GCWhrK7alDujJEsmBPuFvj8eeG666YAcVO+AJmAU5cIfyLXMn+toz9h/VspHSXgK99yBe+7ArY58gYtuSrq6wOT2TLJSxLG6wMt5W4lERiMexPNcvOI6T9zmofb89mu08cY9aHsPo7ephS0aTZ3xB3AxzN1+jPpSmkIyd35AqpOFOwtaK0OwaS5aNEq1CH1tW1FhCKkn2HO8djzdSUAE9WYJlgbQsc7/HUffgFoNaKglP5hSsAHz1YG3LxZpKwZVoFiBK6Isard+3tfydsyVlqxWEVRSi9ulXgVYAljZxtW8/6t6sEWUfKcSaDf0eLJ96NS+41QZmBEmWXkYrg0IVM+jPHHp9kySKbL2UaKxB13/HsCJs/CxwgjO+AethbgtI7d5MFnyya2/T7FPQNoNH/zRJq2o9K3GLbrbctNWiJrTcH3d+6r2+jlCSwYm3T5jAq8fT20ScNstem+HCvgF1kslj9VKXK3vV6Qv22iagdbDS1suyl2fgCJ22I2gdixbS3akfUBmDwVbk1V3FFIrAklCmQTnDUvwrrCJiW1cmMLax22nEuJTYDCr6/TfbiRSP95IQo7Cop6rYjqCS7cH4OW85baMrBru9tYYU7CTjyislT/tKcyNe/Cq5LRq4FBityAHowOHKosGuobAn2wdMm8CqwScNw8F2wZv+oGsd7GFo5mIvdYpH3ca+JoozvHG8fQmgfv3WJ7bGshX6miwvRrNQait4M4VE2iHW42r/+qdgadrsfv2oLZ1SO2gk+37+b0j78TsyFRwtw7JQg6Bsk1k8YTNzCauXMSZnV8MYFNHrKBEN+p02m//06RQ1OFRJlk7BdhERRVIqyxCX5mEje//smw6f2Dnlv59ATLadQWNtNSSwKrNMahwlQ1TiGLrxQ4p1ln/yr14IKnj0TKZaejJJX2aIIKY5DoVrTsKFdUjWastaTnFWhph6bDEjkuoCnKmDb9ufNlJQES+Cds32OL9wJ8GHgD/OvC5+v4/pao/8mU/wycUut8TXtwzPBxY7nvTB7SPOe0Kwt77tx2DtXwvWRBX/51dt7bS4ghD6ivH8omOgDGbz09RJDtktsNsIqWji7GujqyQsqn02trx4HLHA4Iz3752Q68lMNUpwEFjRf9TdwQCak9vVYFFrrLhzLvjS0yy8mK+4NPpPs+HR8YkdHPfPwCwlYzX45jvIphz8aFE5hKqB0FmpfIXqplpUtdL/3vxwO06cL2MbGqrcwoGNmESGF14qItKW4XipfS1ZkWFOXtK5Vs0bCCpELR0WXfeDUgIaDoDhK+OLzsJqOrPAR8GEBEPfBL4a8C/AvwFVf3BN+UZPoHwzzyD7LZ1nZiBgW4V3FwBwPZLX06NRR2ybTcLvSoQwZB3MXKQd7lWAaWyBO2Xe1kVxBtDf/bk7Yn9WBHKVCgbcHuHJkEvj+Xx9Tp2l+EohYt4y71w6D2+q3jAbRm472/JOF5JW7beGH92axsgCGY57il8Lt8jI+zc3N/3QniZX1he4O/t38t9f9v9Ap8N1zzvzXAEpVcTrTzf+rlXIZOs4OlJAOguREv23EpkzqEf/H2tEqY+bjwyCs2TMHSeAGIS7H7g62tYii1EbRVY6NRtm7z84r8NL/zwtzH+Lz/2Jv82ffXHm9UOfC/wMVX9RflqAGByNhehi5E0NfIPhgWstlugRQmKjgrVRVhXA/Bau6BOTUBUhHWJuCFTsiWKYTAgb0lVNAO2pfhyRQtoduji0AyMBWZnWET9qSxr4JWDldX3h3134ZlLZNX1Dg7gpbB1q4FxZagioWPf3NqCNjZseoJJ1i4eyjjuycz74ufZuplBMo/yRMEe46pseFSZiLtX7S4Yq01Ze8zbPPabO9XVYXOySmEtnqt5ZE2e7bhwOcy9wjkdBba+fwort2kw/kQVKrWDvlZHJoCS7NEbIOirnsD8He66FZ/jGG9WEvgDwF85+fcPiMgfAn4c+ONfcWvJh4huRtI2kEbbNViGo17AxoMnbUG7+W/ry9U2EcvRsEazQw+erNhWYn+kCQN18ciJ5t1DoZhCseoQGAu58hJYHesSmGNgjbYGbDtd46RwkwcKwuwD7xge8dDb2q62SMTWiwlREpf+wK5qAF4uZrsdJTHJyqKeQWp5jRF2ohSe9zdESVyVDYvzFeCz7/cOf8Winl2Y+Vy6x0FDb0mKOnzFBwrCi+uOmzTyaJm4XSO5OC5P9gSMdatxqNbqbTwIR0VhW41+6q4EdAtyXyuDXI6CrEa/9s5Wvq2rOQ29poTzHI9vxCwiA/DPAf9tfddfBL4RaxU+Bfy51/m67xeRHxeRH1+ZH/dpfMnhn32ITHa7qjcXobSplYBYAiiBu1qBgk0FmhotFCQercU0uaODkFfimLqefU2+KwVLcbhwYjvmlbhdifcW3CYd8YdgDsZldcxL6MtIrteRmzQaZRatysHQrcHWqt+/LQOfXy+5KlMH/dqfeAJ+7Nzc5cX33cylrNxUktFBY00SmZ1b2Lm5G5EcBUdtSqAdjHRSuC1Ddyi6XkfWak66JKsA2vKSTVy7dZih/40dqN285DZFbtNwFEjVQ3+qtiwnGovmMAy2sryvNC/unAReJ96MSuCfAX5SVT8D0P4GEJG/BPzPr/VFb9cuQhkGiIGyiaStp3irANposAPN2haOVIwg18QQjuNBwwYaJ6CVBHeVbC0BaHGoQoyVC5+MPORrUiil7j131b6svv/Uq/CmOvRuw8qqjpG1H3oT8hRuKyp/6Q9cVgzgoMYWmGTl0h2q4UfsfoOXbmGSwq16DhrNaLTYhKCZjEA99GKH/qCxqw1d9SdojkKlbhbqX1clv0v1BGxz/YMPhMHowm0XQXDHPYUAh2z4weiPhianKss5BXK5mwScNC/Ho/RYy3GH5DnuxpuRBL6Pk1agLSKt//y92G7Cr5xwDo2BXE1FpR7kMjXqnf3VfmH0lD7s9cR1SI6bhWqlAGZkUfLR7qoUu4U0HxltlhSE4gwUbGvMGaFkqQamxdhvVY586mHQFHahIvFFHWv9Ud6WgQt/4KG/6WajDbxb8GwxYDAXu+2NUNQMSZWBzMKxf767syD097UE0F/W/oKVO8tPWjTmZMnmqyjOuP2jz5QglMp+DG6pLYJNBNZi5qLFSV1WWo4sw/4aN0Xm0a+hg4IVRKy9wpf6W/JUxWMlgbqE9LcDf+Tk3f+xiHwYe9k//qqPvf0hgg6RPHlyrAfXn1QCtdfXZiUWKluwaQcqLVj1+DZFkFS5BNnGh6qhbx0uyaGrQ4LtIszJUxZv+wqoBz3YEpHmEZCT59VXV7sNWzSbsKb7P2hkLR7vtTsFx9rzNxXgy3nLu8PLXLpDHyXO6hmksHOFwsLLZeyYwYFI5O5YLSPVi6DuGNC2pdUOZiRVToDd6nKSENMc0KWanLjCOs29ChiqH0KrIhox6tWb7U9pxKcfsZbLbNpzqeSibIlHsztXAq8Tj7uL8AZ49lXv+4OP9YyeZDQqXnDkjSdtII90cpBAlw9TV4fRFoyC3finKEprAyrhyFxFhDJ721xU9xJ04C9CyZ5yE0yUFIFwZBum5Hu7cEdsdPL26fsyroJ5R3R/6xecFD6XLtm6BV9vP48zxWDesnMzL/hr8FfclpEFRxR46AJXRXkxX9jXSOGmjExOuHQHdrJwo0PFHo4jwkMVPzS8obUaG78yuLok1OasaBJkcWhQ8urZL5EpHFepjT51fGAXZnZxhnW0KUiVV+firCXAtjYHXyia661/8uNprZjKuQh4g3iqGIPhfe+BZaVMgXXryJPQbPDVAbmtHKsCImmVQf0TQYLNnzU5WAVZXV0+Wg1GQuvp9ZRkYNVDcuR8Cjwco5eyrrBWGXGIRtNZsu9+Ao1M0xZ1XueJVX0H57ZusfHcSfuwczC5PR7zAwS7zd8fFm515gs58rk8AAtrxQMe+BsbK7qFnSzck5mtS2x15aqY8chVnu6MIhtweMiRyRnNOakd2Db3b0lTVjHqNCYrBownwNFB6SaNJDVb8dEVluwr+aguGan4QPC5j2F9dWYqrQ0A03O8xmt+DounKglQCgRP2kbWnY0Gbd2Y4mdAoURoTqMaKg7Q/lQcQNvYsFGLqxGJxkK4WAkxk7PYrZ8rqWgshDGBKKuKSZKTkGePc4VhTIA5DKdUffPq8tKhOvM2D/6lBILs+56/rVt41tu2ny+kC15cdwDdZqwh+u9wVzz019yWkZ9d3skn3Mx7wsuMknmljPzk/AJbmXkhvNxVh2Z6WmoLIKwVVIyS6lKRI35wVSaKWsWxdQur93cswZDaVjmQxUZ6a/aMmrpY6CYPZqbqMnM57h9oOwxj3dzUBEMNEMwVF2g8lXxaBWRXK7VzOfBa8XQlgZTR+xes97wlgdHOu1vFSEERcxJu7UGR6i1gqkEFNFU+cXJdfowH9YqMhZKFeY2GKVCNRt0RwAoxEaZEOgRYHByENQ92y4fMsgTSHPCxMA1mJlJUWItjG60K2PmFZ+It9/2esboDfy7dw0nhlbzhc8sFzw/XnUTUxoO5Lhy9KaMh/GXgl9N9HvprHvoDkyT+3/ndfGp9hvcOn+eF8IolAslMktm5wgOMg/Bivui3dksShzwSXeL5cMVNGbmPeRE2B2Ba8ivGzGxTFi9qy0uqAetSzUr2Go0QlD0zoY8DwQ79VMeLADrN7BfjIqzd0r2Ku2IxjOUcrxlPVxLwjrIdSJOz3YLBfhlLpP9i9otNjupBN+S7Y8DKCdBqQXaqLRBvxCFxlVCkx3+XbBx/wL7nCQMxX0fY1cThlM12ZqrGGIr90q/FtPkP4i2jpD6aQ+CVtGXWwG0e+IbNS92cs2kDVg1MbmHV0GnCL+ctGek3+KVb+NBgE96DRm50YDrZLLwq/HI21uBV3hhNOdweHYi9KRcXDXgKn1nv9S1FQJ/V66BIqjyJk52Mp1uH5uI7e9AESKGKjY5MQaokORXHknyvBlRd3/x8NGN9Ur9UX/3x1CSB8K4XzMFnCiw7qaCglaZlUHMRzrV3D2rbhoIiQ0a8UmZ3BAYbd6C1BNCZhWXxtWXQ/mkS6tqyaiRSVt+JRTLkuiZLKElwwbgDgpXKbdTYtg+9a3qFCz+z9UbbNe9AX3cH2DLSKCY62jrbRmSYQYFKJ17UM6njXmsVak9/o4FLt/Abxk/y8fQshzLgnXbHoVltXDidKBbXEszYs4OTwit5Yx6Ekrt8GI69uWTpr1tj+NkY0EFVJ44us7jCuvgjB0DurnK3XYqZOYnJBQAAIABJREFUtRq2OFfQE3NSrR4OWtu1vrDgHHfiqUkChAAipF0g7YS8UcpQS8lK/zVBUR0LNhxAbZVVYwf2mVS7YMr/z967htq2belBX2u99zHGnGutvfc55z7qVqW0HlaCVn4Uihr8oYGAFkGIAZH4x/iAMhB/+EstFBQhEIkiiKggiZWAqIGgBskPiwhGiEVRQRHLaD1v1X3UOfe89tprrTnHGP3R/NFa62PMfc6tuveec29VnbU6LPbcc80511xzjd56e3wP2qzKBEoXNtERcmixdf+ZBRUBrciGLyDNNEJsRkLaHItyCV2tOLE1BHnj9p/a0H+9yTKD5+GsAcHZhRKRRIPCRNlcgmOf89/wioSGDMYsweDBDW+G+95w1NcJ2JMqEpVuMuKrgrC01OG+iSqu44oPlquNSSnWDwgCacB5SZiGbM3D2qnITJoRdKg1sHkJYFMPAmB+hA2lBoNxXxrAXPzdntZH1qMJArKuwOEG+Vo1BVuEqgYRVDjExUGTBQBP5Zvl+6QbGtWbA/bCfspY05AyQSZR4RHLGjwAaGBo4Li7OAkIqeL6uCHzAkunxjLpRe4oucUow6c6doafqg6rms/zeAKATuzxpWjBM9YaOrW4mnhnYEFA7cHgrg24orUHOn8d7ynMTScMLmHu33PXY2cyBmq9J9Bk97mJBc+q/IiPwoC5lxFTKAo73lmT9x4DtqnK61iCPhb0+5+SgG+6Hk0QAAA5jsgH1mafIQUxCCRZnk+AxN0oSwDAAkTlbbPr3ZrWNiWrCBrIx1BRumsxYM1BUinyLjEe9DHEzTwMSWXJ7XRzaW4Cet28mrV3FUbFtjlb07r9OmgTsMuE79SHXBPgihdUUeORgSpmUdjtRBU3XFGl4YTUN7xzAvZBpYExi2YhzXwQ/ee4ShBjyxrcUKSXAtDRq1T9YHIJSicWUpMSVrflaLJkd3nsHIDADWMoKKJ4gWaoQFcUUnNXNSPVH0TWgGwQetIa/Lj1OIIAEYgZ5ZDQErqasOsHQjQ9JQOyaIZgBX01LAAJZNihBns5oCUEAOUXGLJQYGSW/QlEgrqylgypIYwVQ6qYhowpFRAUC+8r0NY4W0rEGqPOzU2ee+J8AdF1sZBEqp3u6j6etlcwbljpvl8rL7oWQfIvAHcS8CADbmjGN+qNWpqZEQlc9QfN6MwaGE516GpDIxfc1wkVhA/zEQ9lxO064TQPkDkgLAReVduxmY7CchXVhMQ2v/4MsbJi+wCpi4hok9DVirOblBqCUE1gBdXKuMgVlQLkqSfwseszHwQoRvAf/BHg4Yw2BZQj6TRgP+sHwLNq/HnXn4xA1ANGI0gzBKGls14SdPkxnxZkQqUADIRgEle1GA3WsgdUQsuMTBEhqJbgOUc0syUnYozJbL2KagtepwWfH+56On4TZox8yQ78wNF+ENyE2dh8hLt66ByCifLOn3AzHL2DBoMfireYJeD74m1nIl5RwZErThLwfrvSUqCqy9AxrGZeSj3rSFTxfeMtzk2tyKvN6ntTUNTrkdeAc5pwGjKuhwUMdTO6MtPTU9HxqWdGnl+NsfQSYDQ9gbUoItOzqWgB2aHDT+vj12c+CEgT0LwCzMhXEWUCWtLxX28ki/a8uADCdKEj4BRjT/+FLRB06rGAqnIHEOwxgKahpjbkM+t1DWhD6xeytE0Oi0nwbFpQG+P2PHXd/HNOOK0JUypYm56612HBm/GhjwgrGHM79EbhRAWnNvQOPaClwRfCHVaoPsBDGzuO4IbPAM8YpOG5df6zMAbTD1SYcMbEihd4n+euLTgj4bYckCXozwu1qwt9WK5QbbbfCoMXBhf9TKmS/R0AOgXcP0y4Hlcc46ojvxa70/HAFYthJcTGhKUxQlCBEu8VuE25Kz7t+wciBPBee/lp+frMBwFfkiLqRDrb99Pdl+F//NS/MMqly8eApQNeAGylg+zKAucQWP8gBoW2OmAmFzsVHXbQVHZrgCLdjuOKZPPvbBqF2bjxD3XszcHrMH9E5feGZ9y1yfoG1EVAnT58wzNmpJ49+Jiw04QJmEjlyE8CpemS258BJyF1KzbegmcbS9N0frQSpYIxUkFpocuuKxsTPbhSJVAThDMjnxPOOaFNjGgqQyMXcxvmHixrY1RuOJiMmwcA1zx09iAbnDjZpOUEPJUD32Q9jhypNSBFlB1XQEVC7EJ0uTC/RoxN5BetMLbnVW0E9qBgG75/kj5a5K17vRlj6IlEnk3sNrDXu2uJKDV0fDyTfm9MpbPqigTclsOFYedEGS/Cqc/9XV9QFYdbR/U9yNBP/0Ctawc4FuChMapIb+u94BUveFWBD4MNq9FpsSyEuo7AqQ5YDKewtIT7OuJc02YS4mNU6GeqSk6k7k8zY84RTIqHcHuzSNXgyTYa5M3pmUjUqNQCxV5TQC3btkuAAHzlnySc/uQ/+m1fPp/19TiCQG1oU+y244Ce9tTQpwR++mswsFKBdz0BW1Ro87vfHyz7saFBYpm3UdZeOw/sI8QN/FLNqMQzBkfRlcoIJLgZdVOzjd0q2IQ8U6cUu1nIyBkB0nH8rgWgOoET7pp6FM5GBPK+AAAsXS0IlkkARxJMJBf6AVXYxEkF1ZSEz3Xo/QqfKHijT2wyQLvPGu7sXPVzzVn9DN2iPO2ARoBueh+VutmpW5vvl4jRiJ2WbZ9t+sIZpy88jkv+21mP5hNpk00GPGXfj5UFOiHYBwC6JJ6Rz7ftQu7UYZ9Fv15C8CUP3lltxLr5KRh2vgcB/WFpd2Gf14TVAENXVisXwwYArimolOIqBFUNYAzWJPSxIJN0s1FAvQWB3UgPl3JhFUAiwpEECcBA9LEXyrB7H3tMgronvybt7X2XvAsGAusRaE+llNCnIw3UYdKBWtcPBDSgelDtjkM78JAIddqxSo/tgENPFcFH1me/JyANEEEbGBK0sSfAR3sC+9sWAPoe3k8JfCKwwwv0gNGDgTYJHbWm7LbtImZWfIDfBjZEa26MMVTMJWJZ9c8zxNpn/ueacOC1lwYAdFxHATd87l1/VgslJKq44qXX/lcAEut9E5R8tFqmsFLFkTIWAV5wwERAloZsvIEHS6OUVFSwWhByPcGRs4qZuMTYHsAjWkZRwZZh2bc5A7wSamGs1uirQngoKqjqHgMVG2sQAErg3lQN1gDcxFx3gfe17z2ty/U4MoFaIUxmNoqNJOTHQlM/wn0W0K+X3QkvwbAClhWINxP7A7DBgS8uPlzeNoDQMFSEoGjCJjtlYiEsOYJZcJxWPBsW65Qzlhp7I64ZoMdP/IEqGK2r/tzwjBfh1NWGs8TeO7gxnQD9Ff1UtZMThArBLDbeFOkfQzZXIwB4rzzrfYmlRdxXnTg4TiCg4VQGnFZlVYpjM/aNWGylmRQlEZ1rQqKGg5muMASHqIzKoUuHYfMYuOjkopdTAFDqpivwtD5+ffYzAWIgRrRI3WzEHX/2gB8A2/EvdlHuL9q9vNjr40OfGry2mmwTghCku+F0Zhu2CYGPvgQqItKEcH1Y8MZ0xvPxjEjq4be2gHMdcAgrjrx2aTFAwUB37dAx/2w0Yj+1GU0bg9DyYPMtVAnyK8oYqWIiwV2ruOGAiSJyWwERXFHBAyIeDC3oIiJHXnHHE5YasbSIz6U7TJTxW3ihE46stGmq6FLuVBWHwRkX0xrv9gNaEjhuwD+jbOjAaVxAJJhL6k3U/rkbm7BWZRMGqLDr/nN/Wtt6HJlAYLRkFuBR2X0O+b0gBHkqaY0qzsoFoN48tLrSSUbxdazBTpZ81/TzkyjZKealQWB1HZ5SQQyqnFMNLHQYMm7GBWMsnRYcDUG3tNDlvY+84CbMmDibToCTdwpu+IyHNuCuHtAsQOgEYMMIeFMwGYT43XbELISJCEcaEBFwTQnPrV65awPeL9edhHRqIz4sR7zKk1qSc7UJgXoe1MaohfumB7wvAIRFg4DeJmBhPCwD3puv8DIfEE2j0AOgb/ZDUtmy/ue1yYpLi9eqQcCdoptQ5248rY+ubykIENFfIqJvENH/vbvvTSL6WSL6Zfv3DbufiOg/IaJfIaL/i4j+we/Wm/+WF3OXC3MNwTYqWUhrVX0Y+fzfQEAtCdrY9HmuJQhsuoMBSjqi7Tndr8CAQK1tCje5hn4i7VPaUhlLVgy/X8yHlPumB9SM4xAyjjFjZO28q7SYJnOreQ5kifhiusWLcMKzMONZULsyr9/XXRPvilSR6IZnDDZGVEciMda0BS8KCCDctYSv5LfwXnmGd/JzEyZJuMsTTiWBIfhcvMfzcMIHJjpylRawy6az0reFdOP750YF4FU//2a6Cd4bKG0zHXF/gSqkwKk8YDGEYOLWbcgVoIXuUgSYAGllvFY5PC1865nAzwD4ydfu+7cA/E0R+TEAf9P+D6gPwY/Z109BzUh+V5cERgvUgUC0q+kBgDKBF0KcCfGsp78+D1t/wOnF2GUCnaHmGQRtjUKBagfUgFoCSvGUVpGCRDoOHGK9IA4xCQ4p9/m3o958bl6Eca7pQtt/bgm39YCAhptwVnMSanhVJ7ydn+POrMOe8YyBKt4M93grPOBo6MBTG/EgGxdg9jpfCj5sZ7xTz3i7Au+3Y/cbAIC7OuHDcsRD1fLgKi74/vQhbsK58xjWFhWy65MUG4j0vwPsdtXmYM6aDZXGve8Rd9bl++UNwWy4imry7p72e+AFrE8Q2mUP52kB+BaDgIj8LQAfvHb3nwDwl+32Xwbwz+zu/yui6+cAvCCiL30ab/Y7WcSkZiORNs3PXS+A6nYxaq1qPYJd/bg/4TukGNj6AR1kZP+6FqH1GAToGQEMLOTAobrrYPuEwDe+quzUrrnnCj3ra756TvJxld+BVOEnkHRJct+Ur9pkBiPR9AOMAQjBZFJiDYSXDfigrbhtgpMQ7gxUtJrtmLsQPdSxG6Z6I++uHtQUtR6UHdh4GwtWzcS8H9CzMDGswBLxsA54yCPWFrDWgNkIQ3tMRW2Mpar6UBVC2ZmauB5D/8yBrtkoex/0pwXgk/UEvrgzGXkbwBft9g8A+MrucV+1+37XlqSwSYH1O+30KVYOeHq+bxi+fvhYo7B/kwwy/HoweC2IkL2wm27wa01EnXFLR77pyWaWXLs3EU3my2vlbI5BbDh+NQVxGXJF9N3wjLfCPa7MsvzUxo4QVJSglgCddgxRdGCLONmkwMVFX9aj6g+YuGmWgPuiWUDk2lWM3BXprkwqC1ZoAwXtcBU+FegYDLNld+Xg1ZyL94ChHghs4/degFxOBfxzveQQfPNr5DGvT2U6ICJC3+YQloh+ClouYMLx03gb33RJVA0B2W1ubw46oUW57oCMph3oG9WRgN4H2G96vfI+Ggz8h7z+PqxBxWabHVgwxNeMPUxrz733Bt4ufM0OKkauGA1f30QBQmpEuvQT2pcLhVYoHmC1HkK2E92DR5aAbFyDZIEAcLchQx8KI0vELBETFbwqSlV+MZx7FuBApNyi1vYlqvFKoa7A3gPvDoehzVgARWv/MRZDR1a0kLveAKCb29N8z7YCC0pVfIG6PlFHbBIJGEDZozqfVl+fJAi845Zjlu5/w+7/GoAf3D3uD9h9F+t76UVIVbaGEG+TQEf+kWvhN2saDnLhSWg39B/nBgCAEYH8dbtZCQnIhEUcX8S7DIBZdQJiqD1mVBsRHlLuKjp7dV0AqrZDrY8Hn8dTHxE6dDhR0cBAmhE0sJ7gBuBRY9EChjoRu6UYoCPGhKokIlQkL0/Q8ILP+L70El/LbyLXoKl+G/D54V4lz7hiNCfjpaVugdYEHWbdcQEF1p8RcAE06fEpzPb7DqHaWDT2E9+zBBFCDGo9TkQXk5jW1CC2VJ20pFAt8D4pDn/c+iTlwF8H8Kft9p8G8D/u7v8XbErwRwDc7sqG35UlrhcO9M3ufgOSLBNgoFwJ6iSbZqB/ARebm4IoCSg2IDVVJvLAEO32a5mANOpqN54FJNPRj9wwhooxFoyxYIoFz8YZ12kBk2CuEUuNJtqpevzdEdi8Bk9mBuJWY+8X1RU40tLBQgA6enDijAcZ8LIdcGd9AsB0BVrASMANMRhAhm6yt/gBPzq8gz80fR0AcBNnfGG4w8gFIxUcw4L38o1ak5FmK0zQCYorOdtH47e9vdFJWmEb65XGGLjgOurnkCubu3HsrMExlZ3WoJ76tVoggPcFnuqA3259S5kAEf03AP4ogM8R0VcB/LsA/jyAv0pE/wqA3wDwz9nD/waAPw7gV6AMzn/pU37P3/YiEbRo04FKZhhCHRBUrQRow+6Uv1jexrZ/bObcKrq/AAANFOZhCFI2G4emF3W1FCRsNmPuv1eEccoJ10FtuIB4UQIMXHGTZlyHBcUkxs41oUUTCbEuvzf53oz3CBCFBpsvAUMtxR7aiM/HV3iTMl7wbPW+XgY3XYGIsIhChgHNUk4t4e36HC/rEYkqnodz9z24rRujMVHFfR3xuXiHIgGnJYHmgDATwqywYWoAL0AbtFkbVumqz+GqYEwZz9JGkz6VobsZv24p5gCrBkUHtsaIsaEAkKYmLlUIOacnsNA3Wd9SEBCRf/6bfOuPfcxjBcCf/SRv6lNdxKiHpKeMN6X8H5sU1IPoWLD7CVjjSlw6nLoeIYWGlrmrA3XAkSMJR8sUSFALqwFnISB44KCOZquNMKbQBTHu1wEtXZYBTQjFhDdfyYSlRXx+uOuqvICOCN07IEvAW+EeA+cO5mE0DMYhcAHQuzYh8BkAOta/ScZzzsjQhuBIgnfqgPfrFe7aAS/rUfEB1LrpyakMXVDEZcwaqBuiMuvn6VwBrooJCKuogEtC79hT1WzpkApuklqoP5QBkStScO8C6sEAQE/3vTzwfkEpDA7NUIMbZPtJcPSj67MPGwbQxtAxAh2z7l1ph6iOsrED9yhAoa3hVwkSHd9qm9+zikLab1hZv0iAZEeW18QESAmQqBdnDdt4sNnskl/DwjdRwZFXZcR1XLE07by/YaYfvAP5ZMSL1F/re3UjXiXgTZsSOMDI14CKldCNRV1d6IMW8ZXypiEMVTfgzag2ZktLOPKK0R67tIT38g0a1IJsbgnnmrAsUXUbPUi2/WhQ1PuRgDpYeVZUWam0gLsyYq2xfw4ANk0Gkr6fAykhS/sG9sezjMHdnzQIPEWAj1uf/SDAKmpJTU8hhfXu8kkLChLMhoxlYwhevI4owKjYBV03BCEJXbxkpym6JqGNFsmUjJuQZhl2W803qOMENg2CcGHeca6pC2ooiUhHfPr/iNEwAXNLmJCREXopMJDqCbbGF5qEE1UkamDZAs9tSxip4mWb8OX1c2imHQDoa9zVqfMGKhi35YgGMr3D0jOQJmwYAZ/A6OfQIlTlhwwz0ICwAsNLRi4J53HEg1mzAcBDGTCXeMEg9L5BLqF7EwZuqExoJWyZAsQsy82x+CkOfGR95oOA5ILxNz7Aq7/3+7a03ScDgu4j6AKivZ3fy4VdUGgE5O21e5mweyzcfgzYJgZJwLGB3N8AuDjJmAQptF4GJNbMoECnD5GbafErh2Dg0uf0Fdxpwa4L0Hb9Xi8FKjbZcVcfutQAMC8Ckt4j+Hp+A792/jyuw4IvDPoeTm3AqQ2YOOO26mh3kaiNQTcjEcYiAffZzR7R6cMtAMzosGH/O3DWi7FFYF0Zbz88U+FRy5J8o+8DQWuERr7Zt1GqNwUJW+YASDc/eVqX67MPomwV9Ve/vIGF9rNp+9elx6mQwYldiHQXHGQXCByWaoxD7EsGgk4QUlP1oKABwMeF7kbkFtqBVXvHIcIAuqaeE4dcHKOBMJeE0kLvCbiop2IEBFliDwaA0osdKxAgYLRuY+5yYz4dWOH6ABEPkvBryxfw5Ye38KqoEpGbiwDah/igXOG2qIqxux65lFkWRfrBPh5ncG4MTOmcAd5nCqSf8d2izcAeAGivFSCdgh0Mc7EPAK1pabWncQN4CgLfZH3mMwFfLWx7tV8HfQNLTxOpmXz4jggEgVpnjd5UsOc7gGh/AdtYsVuQ2X0uPe54ARXCVOJLM4y7n3oFWnIkNpdem5NPoShYKOipOxkqcJakp6lt/oGqbdYI0GpYAHRL8RUBE1yizIxK0BBAGFC7Gemvnz+Ht+9v8Nb40DUEggWe+zoafBjd/yCgIZDy/6uJiohDs/cYJodcrDvx0UDbZ8iCaTf62y+fDETb/DFsJZZ/9YkN6ec8xNpHik/lwEfXowkCEmibU7tQaP+mfglDA0BqvRHoYCKqlmb3HB6XBKJGeoUGAUdDBXax0W3zB94kx1OoOMSMBu0PzDkickAK9SLtjdwwt4QiAQMXxJ3hyN5tKJiZ57Pwqp/Y7uuX24AqjFVGvKxXSFS6HqAKj5QdrbjgG/kN/PLt5/Hy1RF3z0ZUKClppNalzBNV3IQZN+GsmYWoxJm/rm9Iso3eQZjd+RkbmctLNFNhTqxoQacSa0aQ7PNU0ZVpyEihYrDPqysOs053wk7nMXlAeQoCH1mPKAgYGMUvAhsPdgtyhhmQYvsqBF4ZVKCGJZY1ULEX4Q0ToEex9I0P4GKe7fThNUfVCxhXHFPGGAvOJeF+Hk0iu3bxjLWGDhwauKDsanhNv11LUJWEBhMHcV2BRAV3bcIHct0xAi4FduTF0IMVz3jGDc8Azni33uD9eo1Agg9PBxALjnHtBKSJMj4o18gSMHHu0wHfgA8WlI5hweemB/yyjVu1Caij2DaopiI1C5L7Ukp0wjKXCIwaSOa6NQU7RTgqA3Pv2+B9A7V1Q4dn+0ixVkasT/XA6+vRBIE6aCBwFqE3pMSMSREbUIMZjGxza8omiS1Qs1LSRuJeYARBQEXBQBQE0YxD9rJiLnDhq+sNQjkCN9OC0hiHmLHUiLIOvZZ1UYzrsCByxbkOva6fjQIMKBqQSbOGDGXgAein82099pn+/jk3POPG5vpHXvCqTfh6fgPzkjYWHgmC/SdxwZu8kY5ehBNesBqhfq28gbklnDCAqeHZzRkv1wC5i4hnzca6Vbif/gFAk02QJQrWojoCEcBSI045YdmpAwVDELbG4FEwxWIKw4ycN2hxrQwKW79FnsaEH1mPJgj05l9v/NFlndoNLLE1ASv1epYq0A7Sx4I9e3Bl4rBlAk5l9Xk2gH7xekAINgFY28af95FYNW89Z72WxjjJAETtvo9ccVsPfUzXab1tBDD28aGrDzmIqBoJ6INyhc9Nr/D5+Kqf8FUIGfq4d8sz/bn5EmufJYCp4bYcceQVb8Z7POMz5pbwtfZGd0l2/4NvjM/wtasT7u4PaA8BQuo6FGcrrwYCFfubBEIdLVtbeRuTCu/GfeifKwCUGhBNhBRAzwQAoJnGQIgV0SYvRMDtH5lRjv8Yvv8//Nvf1uXzWV6PJwjASoJkXoPVmlJJOuBHH7R9bdqDdgFaUOiAoz0/wMRF+88SVxmmPg3Qk2kTIPVGlivn7DX0X4fHApv+HhtfQDdduZAgnyXhVEfc11GJPVR7536RiPsy4q30gGdh7l4EgQWZWncX9hVCQ4gN1xZ8sgQcoXJmTA0vwgmTeSFWEGazQVNAUjDugIBDRU2bl0OLOiYENFsS6N+Dswm+Qje4G7F6cBRsk4FpyL1E8I2veAsvGzaUoEuSEwnSWFCuvqXL5dGsRxMEfH96SdCbUh+nFgzpAKD9Pu9dbh95XXwTnXdA1iNwhJpnoEyCZrXrpU6Aeg0wSffW20tk+9RgrglMgqu4YGkRpzZY488BRBFLSzi1AfdVM4I34gkBgttyxG054AvpDs+jpu4eQNgCgBuT+Il+OKwYYsVVWDBQMRcj1yuo5nfAWBH66/liUlbjIWaMY0GZquoKCKNOAl4JXDckoDDMF0Kf7/ePRrXOTU1YhOVCXwCAUbRb/xMCAHGD2Lhwz3J3O7Snta1HEwQ+wvUnsj2/ywTcHadtAWBfMuwzg4vlWUTYAEBE+MhJ7qfU3mWIbYx1QO4XfgoVh7TdBvRCz8KYq8JxDyFjNugugC7ocVsOeFUOaCAcKHdK79JUP+CL6RaBWuf9T5QxoGEF41Wb8KpO2vSjjOeHGSkYGxCtjyQnCwhzS5hC7jDkibWBWM2q7MgrnqUZV+OK9RCxNkIVAMLgCcDsI1TL0gZrwBqTsDTGMa5ABDhrluEGo7Vxv91LgB3ByEez/tltoKGn9fp6PEEAuDwBdh3pC0CQoDsOu9gF7NvUCC3uThUhCHYTBfvH638HrYj1CEphU7pBr2M9xXfpMA8MiWuXGAPQMwR13/l4XvypDXh3vcFdGfHmcOrW5VkCjmHFtY3zABhiMPaf2Rp3X0LHBBzTupP/1onCxBp8nIg0sduSbyzCAEGDmpFcxQU344K1BIgAuY4KkS6qOuwAIs0EVNwVsSkD0zMhSEf/eUpXGyFEpWAnrsjWW+lOQ4ILGTe/T4Q+Th3+Ua9HEwS8BLjQCAS2DbwbHZIbZfgTSbZKwfsBjih0oJGQKgwXRmHepKxsgwFbdlAqY4jOG2CI9QleT3OPccVNUsz++7jCuSQMRjf2k/+uqrpPoqrCn+vB6McKKFIAT8MUM26CCo16Dc/2GoBzAg64rxOeBy0X2E7cl/mIh2HEW+He/Ae1LLjhM65oRSWy8WPs8ONZEpamgqhXcUU56P2vGqM2QM6xm8F0fYekPRpODW8ez51OvJJgjKVLspemOsiHVDQrENUZcCs3ZkEtGsrSUBR3YX+817Ozp/WIgsDriLW+yRmK6S/eAKCLoOCEIlcYdlSgJAEGhQZLYSDThhPYNfU46KkmojZbwDZTBxSBN4Sq+oEWCBiClQMGVk1BJp3VR6r2WNu8DVgQEUhT5YcyYm0Rx7jiOiy9hldosY4F55bwIp4uRESaaCngluYnwxO8WiZUISwtGu5AMQV37YDZyEt7ObNkZYJjFL4/fYj5kDpgyU/220aop6A8ApN3kwjUQT/jlDQCPRJFAAAgAElEQVTQNftbzHWTFvM1xKqBWUj7DkF7B9qQTWiWASSbDtQSNAtjwVNVcLkeTRAIi+MDaDvBoVgAKaQOQ85/iTr+E1cCEujJH6UjDkW8DCCgKJVYEgCBzqa9Jm2EGIEQai8TksmK7ZuDfvsm6sk/mnpvpNqdemdJOJXQpcYLW3POUvFzTbiOiiUItHEEAlRmbKKMN81n4P18rUKj1uRzxeIAwW094LYcuunn0gJu6xFMgonXTldmoyk7XDmh4gXPmCXgrh7wdnmOUxvwLM4IB+nIvrVEnN4itHNArQrIchwGBp2kpFDxPJ2xNp1o7DUEAomCjQBcD/peHvJgsuPmTxgaxqF0S3fCk7TYN1uPJghw1lm+eOruZcEuPeyb378XoQV8wQYL9u8botBlxJ344mKiPcjYqdMao+QIDlXlsVPpza0mm+fA0gIOIePgdS4REjU8S3MfuQHAIazdO7A0NokyPQ2v4oLPRU3dPUgklI4SfGnW5FdmLT6g4orWrkicJeA6LDikjLlElHbpOjxRxjGYvyFClzkfjH7seodKWa44ixqTHELGFDPGlFEPjJUE7RTRDlUzsqEiDRVjKojUsDaVEfOgp1yB2surfUY1cMVKoTdaa2WsWS/vMRXEoJkDM6E+NQUu1qMJAmGRPt5zyLA4TLhwFxBVVSFrJCXTHLRrjQqZGrFtcqcgO3fAygFFxFGXF2duKMVNMDSVdessAJhCxk1akFvAqQwIVtN3RWGoTbdjBFxiLNkp7r2ESBVDKngez2oISg0J+hq+SRNVNSa1Tv7RAsEsmrY7lPiuTojm7BOtaek9jIkXPLTRAoaanA57YBIUvzBZU1K/dONepwX5EMwujCAHgmTlZBALYlL05BAKqo1F1xb6FKBag7Q17rDh0hgNmhnsgUG+RAgpNJUob/TxehGPeD2OICCCL/yv7+DVj35RT4FdExCAAn3MUkwfj940aNEmBI4QBJRgxLvGgomSoBmDzUwvOLXLRhRtJCJfDhjqKX5TV96BS998ziJkUinwFaoxOKYzruOCo3XovWF3HWYAsA2be7o/Ue4OQgkbwChDdQUHqmikfIDbcgDDxDos+3DfgmyWZ7OkzWlIAmBSZg3cuQoA+vtu1DCFgpu04FWYNEDGptQCCwB7N6a5JhTLlsZQzLJNG36ODsxNy6PSGK+f79voEL0YeJoOfHQ9jiAAoP7yr4HzF7UD/dr3qGrjrzsPM+mVAwKCwEhxW/bAqiMIEUi1yLCTJNuLj74OVWfTEGDSCzfXYLN/PbFVT9BdgaxMgJgdl7LquAqWZqkubQIji8mD6XMFwRpygRrYREVW/2WgQeKlHLv8mHoRFHM03tyAoysa1xEv6diBQgHa9KwgVEm977DXM/DGodqoKxPxOi24GlYspgBE0AbqOGaMqWhDtCMquTdC264nwLzRsIso2WotEaVyx2NsOAIzLMVHsRtP61sQFfkmZqR/gYj+XzMc/e+J6IXd/0NEdCai/9O+/ovv5pv/dpd6BW7uuBcegs3swfbKQMBHpgKe8hPtMgpAn7NTKnZFYsClrXxWTRcCGW6yWYS7gMjHLYaepkdeMYZy0VT05TLfenszJ/G11xZ0R+L3yzXezs/xbnmGuzop+cgQgw/r0G3SGwinNiBLxF2bsHS6sJKTnD7MPpKkjCteMJIj/gIa9Hd/lma8GM+4mRZMY0YaCoahYIwVU/yojkCDcizclZi5IRn2ojs22YiwGEpQRHsGXkaUqm7FT+uj61vJBH4GwH8K4K/s7vtZAD8tIoWI/gMAPw3g37Tv/aqI/MSn+i4/paWmF7vN00FC2vF3IRDZgYU6OQjY6K6uE+A6gy4gEjZjPTLlW50IbCcToBdnok0l19GDrxtvqgqw2ZJzxci5YwC8SbhIBOqEQA3Pw/lCV8Ddhlgarrh0u7G5qWfhy3rEO+U5cotIFjxObbTNHvCwKJPxroyYjZCUqCC3EbDx40MbcWojRs54gdOuDxA7b8HVhrTPqiPQZ8NZa32gBz4PjLUx5pLMPIWwlHjBCQiETZzVPs+9+nBXFrZVLEiUEiDfJMg+5vU7BgER+VtE9EOv3fc/7/77cwD+2U/3bX13Vm8MNtrUgIJt4JX6aIlY0AHAXWKMNj3B15f3B4RAsVlG4RgB6fj1NBRMg6nzlnjBeltrQIxeJqjTkI72gop1CGNkbba9KhOurBdwHeYOCqqWjrttOVMDGDhi6eKiWSKaWZLNMmBpKhyKit47ONUBS4s26pSu+Bt2uINZIj40Jk6iiuf0gGc844pWvJQDHtqAUxsxUcEb6YSRSx9lAjrOnELGPY2ojW1z76YApqq0lIiHPGApsX+Pd1mUQINGqXyBJWAzdkmhGeXYLOI/hevos7Y+jZ7Avwzgv9v9/4eJ6P8A8ArAvyMi/9vHPel76UXoK56BtZChAu0kj4I2NIQzo50i5CqDYrMGn2sFCtp90jHWWEEM1CUAi3a1ecogFrTCXUMQgKncUqcPx6Bp7HlNVqcCKeho7/sOd8gt4CUOmExt6GysvEQNMWQkrl1hyKcF120BWOG9Ey048oKX9UrBPIYVaMy9/r+rh25XNhiBaOKM0br5p6YBYGkRyzkhRD2NvT+h0uZqQ+ZjwyOvmGXAq6aB56GNlhFo6QAA12HBMSwdh3COCbd56jqLgKoJHWJWEZNQcCpaYkRuWElQKiMwMFjJ4NlB3WUQ3YMQ3tohy7b0GhBTiXpa2/pEQYCI/m3oFP2/trt+C8DfIyLvE9E/BOB/IKIfF5FXrz/3e+lF6CuelbverManRpCKDlZB0I3cOQTBywO7gqI2ApubFthEoGXWSUAl1DVCxoo0FkSXGLdAcD0tXU9QpGFMajkWueHlqqSfKWQwBEuNWvtDx4JHXjFSQSNVG365HjBysY3ecGoDGqmhaBMdD84tYeSMCuUFOL6f0ZBQccMzXvIRd3XCSQZrIAqKBDyUEW0JkEY4l4SlRaMm64k+csZcUrdBW1rCu+UZPqjXmFvCTZj7yV+FkS2tZ2hTs5mhimMbauMuGlok4GSExAbCYrRiHwmuNSCvCclKrvOa7PsmLFJ1UrAadNjFRQBDDD7hhi7WdxwEiOhfBPBPA/hj5joEEVkALHb77xDRrwL4gwB+4ZO/1U++hjulsKruvezUgrFtfAAo2ihEbOChIQ0Fc2XVE4jKsRdRrwA02p7rJiM7NiBzU2feXWc7clP8ewk4pKzd8rjioQwoLeAYV1wZcnBpEQnoBKMP8xEPZYQ7FFdh3NajGZI84E26x4twQjWBEBUT8dpZgUPvlmd4K953rMDE2RqPmmk4sCcctiNzaQkfFsab8apLimn3PnSmIwCsErsfYpaA3KLhFaQHJhU9iV092Tc5GkNYgxyT4D6PfUR4zrEHCnUg1jrf8Ra1cYdnt4+p+1WHAKhoT7Dh19Z3FASI6CcB/BsA/gkROe3u/zyAD0SkEtGPAPgxAL/2qbzTT2GlhwZeInC0TEAUDShRNgrxxywisRKBEVNFTBUlB1SEHjg8MMhJN3wb9ChjMpYegCmqe+/Doml+YJ1/X4UVh7Diw/VgAiNss/XLvHXijOu44GU+6mmPyy760hS84zh/v+/URrwIJ+3aIyOwpeyk9f11mHtqP7GWIkV0UzmOIQuDRTMO9y9gks5HYGzuxktLuA5qI3ZfFSuQuIJFcGoDHuqIpYUumKqwXvsdagTl0ZSVG6Kl/UOsqE1LqGInvfcQLhyJbHwoQiiFEYJgSEUxYcU1z5/Wfv2OQeCbmJH+NIARwM+acMbPicifAfCPA/j3iShD22h/RkQ++C699297DbcFIUdFjBkLkAprf8CsybqvgCkFeT3PQeBbMkXVAyhrULSbLTK2IZrKW1VRdFq1FLUJ4TotOE06enO8QANtcuONMVeF6vbvG1JPmX+aARwAHELuXX2HylQhzDJ2AFIDwczKDCug2cDckopyWl9gP0qMVMGvXRpNGMUkyvyEV/Rh6FoF+WPy7E7zFUbF5rjURKHOKVTF/Nv9/jlGbt2ObW884mAhf1zgrUnIJH1SoCXYd36tPKb1rUwHPs6M9C9+k8f+NQB/7ZO+qe/WGr76IeKPfwn0ht3hcOCsYCG3BgO2fsH+gnPLcd+ceSzIPnve4wYaoRXSkyeqjkCrCmiJ3PDmdFIFXdt4uQWzHGubfoBJ7Tjc1nUDTy7hFZfuRpyb6wIY0s7w/9ov0DW31Md3d3XSYNC4v65v4E2q3KDRjpwUwlIVKRik7V5fJxiuYsw7s1JAJweVts49k2DkgpUjinAHE+1HfL4cYLRpL1Jv9Hn/wL0HfDoguy+HbAs2UxJp/Lpr/KNfjwYxCAD1V34d0z/8RZy/4AaZeqFT1RO7TGYVBk39qRLqElBiBXEDR0IwzEAMDeNYFAO/BFMWdkASIJlRcriQvD6vqWcDkSvWFvsJvrSISM0ow8oPWBGx2HguuZhoGVGaEny8Iff6CazmIJvOQBOd5/tc3zv2yw7kkzqvIOBcVbYsxoZl1tq7QU/q+zpql99QfJ7me4PQRUUAmJtRuAgAiTICVOnooWhAI2yjP/cbAC4NRfY4AbFgvS+12i6A63RGH+/BpTW1h2/NxE5/5IdQfu3Ln/yi+gysR4ecmD6sCDM0CJjQqFOEAYP87nkCdUsr2xpQyyZgwaT69wBUxKLtEIaNUJaIZR7QbN7uisJpf5EbGg6AcujtvnNNKj0ujLWpK/HLfMDLrAzAyK1nAn56T1QMurt27UD/99RGvFtucF8nnOqI3CJuywEv8xHnugUSDQKbyUfL3Jl8TRjnOiC3aCWBypffGFYhsesQ0tY3MAjyaCInIynQ6RByn/enUJHMjg3QDABAB095SdRl2u2xTdBHjMH6AD4OvMgo2i5bA5B/cMHf/de/+Amuos/WelSZAACM75wwfvgM63NGMbuxlqw5JQTJpJ9KbJCgmILlYUCcMnioSlFdI1ZEJQOFhuGYUSsjRIEci44Uja4cYsU0ZlxPS3fBWepGCHIxEd94Ph9XDYFNRKS0AISMZ2nGUuNrJ7c+98irugcTekrubsK37YjbesB1mBUQ1DbXX0cm+n2JKw4hK9pxUUw+AIxGanIRU0AziKPZnVdQ1zxsQj1LCSY6qk5IseMOGNIVljPtfQe3Xkmkhkp8gaQUITNqacj1MgtqbesJiADjqIYu85ogjSGNIIXB5alh4OvRBYHw3i0OH9xgeYMhkVGvG2SqwH0ALZswyH5SIJVQS0BIdoJbd5pZue9EQFkDeKyIJChLBKAMuWEouJ4WDKFirQFvPzzDi+msCsM2CitNG3VOrplCwX0JXV3YJcaaMJ7FEw52+laQoftSHxee2tjT8Mk8CbypmFvAYlZeS4s4twG5Kc23tICRs3bjRf0Pa9WyaS0BdzaWHEPpQqdvxBOORj0GgNzGjlR0/sGNMRrVoHTCuSZVTyIdQxbhrSR6LTh6X2Ev0Z7NwUnJQQWJG+YcNTvzdN/GtmkoPQPbqzc/NQUu16MLAu3d93D4xhdw+sKE/IzQVoIcBG0UhHuGzAw51g074IChwioZ4DNxUQ+BXIM2nypjPQ0Yr1aE4+ZfLgYOej7OuEorXi0TziXhKm0n5lyTPeZ8+V5tMy614iquYGr4YL3CGEoH5OiYjpErIwzalLuvk5GJVPnX9f6cefg8nnEM60UvoUHrfVTgVAYcQtbSyFJur9MB4FmcNQMICz4o1zrvb6mXFjdp7gzAvayYax/4+ziEvLELm8qpedNVn2OMSxuZujCIW4+XGlAq4+E0ohY75d2ROLQLYZdSVMPA/65PcWBbjy8ILAvSBycMr0bMbxFaZNW2Sw0SGGFmVAYkNbUmA4CF9f+D6t/VKig5dnbgEAs4NdSzbqohaW1fip5OIooEBHSE10pE4orBRUOs877WiOdp7idgFcZagRapi4k0GxMuLWLi0DdbMXchJ+u46lDD1v3fp9QuO+bvKey+d4yrMh1ZQJmx5Nhtxg8hgKPiAz7MVzi1AeeasDbVOJhLstO+4lSGrpnogih7wRQnE+2nIoCWZt403fMBHAjkPIOlKCmo5ABZQqdR62MZNajJBLFollCpB7YnEsG2Hl0QgAj49gGH928wv5VQD9Bs4NhU6fZECCdGuRYgyU5TkExAdBs9+bjwasg4DQX1PqJWwmBIu7MBVpoAx6jY/3dON5tHoG3gZiOvc0koRhQCgHPWcmAKBZ8bHtCIcS6aigPo0t+Jq4qOSkQF48hrBw7NEnFvFOEmhEWiwpCpYg7eELwkKTVW4dOUKtYG5DVibRGRKu7KiENYUST0YHRXRswlbZZqpgGwGvpx4Iq5bq9xbeXNhew6NmKQN1Cn1yjTnWVotOBSGDWHjc3pAcNBRJVQWjC1J7LH0VMAeG09viAAAKVieJkxvR+xvCDQoM0iQLXwkYE2MNrY0H0IGtBceHQ3Q/cadhoyZkyQxhjMNWctAQUKcCnCOFBW6mxlLDX208+7+2sLeCiDSnQbgKY0nQ40aMNOT3xH5inRp1ozr1pA6ZRlUB8DLs21AlmBPkFxBjkELE2Zi2wndO/Ox4pVdupHLFirjhCbUGf6+XtczD24WPrO0BMdQC95CitIqJmQqr4v7j+z7T5TYHNfAtCNRWrV1L8WRlt2LrNG/YYJv/h9XQfSA0Cjzfz0aT3SIDAk8FJweL/g/gcSyrVSiWHGo5yBOhHawerLQpDkF9TOZFQI6xpxGhKmIYPG1uGrnZJgM+6XywGRWr9vNiqxdv65A2C8Gbifi3u54GIizTgAuvmUjOPGpK7+O5sHQJbQH6dgH96APcYXAGPzLaRqm9wcfqp5KTQGRw0WPsYMJABXDFxxggl8mC+AC5/qqZ42HUChTk1myxb2luub36JgGxrqa7uSsOxAQ/BJjNPEg4CiXIK39ie/BYMnJuG2HmUQkBRBWbOBdJ+wvKUiou4wzAXgFaCFN7yAnSwlB4QIBQ2J1prnNeGN4xmHmxk5B5MNY+Qcu9TYXBI+WI4b/dUUcRpTH3M5a3A2Cu0eG+/NNIbO3EtToFAV7RdEg+fe1oNpBB43FCA08zjXhDGUTixaTCuwy5lxQzLxkvuq0wCuQH2I+OB8xHVaeqOv6yKSvu+44xm4rNch5v74QsHUgfTfIZQeUJwp2cBoRF030IFBdefYTNargKf2DAUMEG1BIDSQ/d2k6TTnddbQU2NwW48yCNRf+lWEz70F/qEv4ertiocfNIkhAdogqDZDpkJoU4MctD9AwSGohJQqeGjIOeB8HnAYMj5/84D37q9QGyOFhiEVrIgQAeYSkXejqly2069URgwNz8aGIlr3R2q4SmuXHbvPozUIpTPw3JpsnzLflgOYph40nFrspKCDHYt3dcJ9HXuzMbfQexFLSwpjDg1oQLwL+OD2CleD1vfXaVHhUGiGwhBcxwVl0PfuY803xhMiNTyUwd6nZh+5Bbw/X+EQ84WgatkhAwENnN2ivDJyCSiF1Wi0sPI2Kl3uaAGId2Ahfq1UIL/xtHw9yiAAAKgNfDfj+FsR6faIfKOXRp0EdbR6kqAnzdAQpwLihrJGpbsOFUNUjMD5FJBLwPe/eA9ziXh1mnB9WHA9LZhDxVoi1hIQA2GKBdWw7i6KKVBGoZcHgAqOHqKKfeioMKDkCVNQf79qp66f7j6rf2e5gVp/LeZCpOXDQxmw1ogcNuNT1/X35uSDKQrdFWX+zfny8hB7Hy9XFYFxLwQAiLw1/HwSAGiZsXAA6vZauSr9uAuJdj+FghY0M1qqBU2o6pJrB7Ya0DJDVgZlMy1hZYRKaqDY8PoiFhsZ6qhXCp6ag7v16GDDvuj5DcpbVwhzwdXXBVyAMGv9X48N9aoq/BdAmArSUBCCQCqhLQHLnLDkqCdTYdzeHvHV+xd4Ps6IsXaHnMFgxcUksM7Z1HJCU/8BbmiNcVoUJvxsUKzAUrRhNnAxroGi427SsjuxtxLB9QMAdBgyk/SRoSr4akPu/XyFLAE3cUYizT68wXiuCfd5xFoDplQgUbOjaO81svIbAOBVnnBXxt50PJUBH84HvJwPOJUBd3nEg/UXIjdMIXerce8bAOjNxX2/wXsr5xyx7MaxLvaK1BSibRBvSe1C6LX3ZJo2BvtokAUYK4yH9bTwiDOB+tWvI5WK5ce+iGe/uWJ+a0S50n4AGHqRVZ8p+1iwYo0R7RRRTrHTVSUz6BTw9fACX3zrFoEEa4kILJhiwRALlqw2ILWhQ16BDQ0noso7a1MMQWXGECqmsGEJip3oz9MZqfv7oafT5zrgVIbekXdpcCYxc89m+P+ERK0Ll1R7bb8YppAx16SsvSRoY0NKWydt4IIZCXONmCyzmGvEuSRN51lRinfrhLkm6yWoVmFXL7Y6f29sUuRSLrwK4bwMfRrQ5cljgxApLMC7/EGNTINpRLbqQUMfT6wBAdY+2LtLP/b1aDMBKQUIjNMXB/BScfV222kKYFMWLox60hN/WRLaEkALg+aANge0sk0V6l3Cq7My9HIOmLPCYA+pdNKLK+PQrunntx9WBdsMQfHu3gQ8hHwBMV5b7FBfYGsgnmvaTlPhTlTyxz2LC54ZGMlTf1Ux1sfNNVqg0JT8MGT1ByR9n4nNEZk2DQQA1m8I3VaNSbpA6Lmk3nj0HkLtJz9voKadnoH//NV8CWplRQTumnvEau5CQwOldhEAAJ/MoE909En2t5enxuB+PdogAADy6h7Xv3nG8taA4zcKjPsCXtgkxgCeCeE2YLkbkZcIZAYvDD4zaAlqPsIAGOAzYz4PuuENqloa45CyimdYM9ChsQB62hstM7jLWo87zz4b9HYKRUduEnBvp/1+hKZW5cYmNFHS0QVDbPNFrloCcFWGYguqQrQ/jVvoG3KKZbNqg5KbBi7a4TdtBA9MXr4EMwWpltnMlt10wZQS+zREesMymKLSJr8OoNuGtcxoWQNuM9CPNK31OQg4Svd/3FJ/6pMZt4YTRwz619MC8MiDQP3wQ8S/+5s4fT4gvVwwvIJu5oVAK+s8uRDSPYNvI2QJ6k/QzFJ7JQ0WpP6E1NBPLDfFXLKm91dDVkmyxhfS2s6J9xJhLpoyO7tubbqJ9k3D+zx2nQFgo+smGxUe44oDryr6sXMAClDBUocrA9q4G7j01wZ0fr+n4moKLRhsFOijvv3zHMQUjGREUBGQagFCywu2r00UpAuH1rA5CtvvU531VzUoS2YNCMYT2I/9yMBb0lj7Bzt5eGmENkfIOQJL0ED+hBPo61EHAQCANJUXi4zr36oGE7ZAkG1C0IB4JiCrrbk7ElGxYFE0G2iT6gicl8EuRMJaND1/Ns5gM8t0TT2df28yWW7d7XNypoa1KoqwOBnHeP2rPWbk2pF7Pq/X+5QufF80YERquDY1otR5+g3HsOI6LN3MZP+Va9Df0U5o3/Bz3azMfArAhucPJhkWd8o/i/ULijBSqAis4iGeRXhGsM9smqEDxbD/yhhiwMaCYupNrRBa3cBDItSRnQAglSHnAL4PCHcB4V6zOF6eMgFfjz4I1Je3ePNnfh6nL02Y3stI98ow45UQZoYEQUsCqkA4m1qNkdG4arlAReXJJDVIZqyz0l2jNdNul0nHYqlubkO7kiCYSal/zSXaia7UXm+oRW52GuvrRq49la/QGXy0mn4yBqHbhbFTil32y59nakAv0ql3/de2QZopW8/Dfp6n9YD2GrwHEaj1zr9zIVpjhU43r/s1UxiC9hbcc9CfUwxyvIdGE9vYz2GYnsYXhswBctYMrZXNYMSzGLJpDj8ExAdCPBHCWf9N95/qZfT7en2nXoT/HhF9bec5+Md33/tpIvoVIvr/iOif+m698U9zERPmNxi8VkzvSXcaoqIbvo3GKj4TeFbDEYmC3suy0RMfigqYGsElBG1WzWtCA+H5YTZFHOobg8mmBaaMEyxL8A6/G4OuJj92E7XTPtekKbZtrrVFfGDz+2gmJa/KhPfWK9zmCc3GgGoxxv1Ud/6Bin+IjRN1M88l6u9pnfSTSaJPIXeC0KnoRCLvxpJNVAzkZlowpdIDXjRzketh0VJg11zcSoUNHThEJTF1f0gPBIU7XJiqlmSyauZFLAi7oEFzQLq3jf8ADHeE578EfP9f+Nvf3Yvq99H6VjKBnwHwkx9z/38sIj9hX38DAIjoHwDwpwD8uD3nPyOi3/NWD1IK3vqLPwcQ4eqdirCQzseTlgYtCqiRXkwPeuq3g2YIHY0W1amYUoMUQlkNCmyGGffroB4DZkACqKzWENWAxDeGn4B364hXedINVkNH4p1rwmrp9akojbeIgmuYBFdxwbOofP6X+YBXNqZbWsBDHfG1+YU9Z29UGkzJSEsJr9OLw21p0xTwjXtlwKCyyy5cdATQHsQhZtwMC66HBcmyljGWLpzirswODAI2O3HAHJyDjifjsYAOBRhaN4ylhREeuJdkXdGJ1IuwLUG/X3b04QZweRoN7Nd35EX426w/AeC/NROSXyeiXwHwjwD437/jd/i9WiJYXySEc8X0foBEUuQgCNU2vMwbilCtzK1/kAm16Ay7N6NM5SYGLQHuZ8XiH1PGOacOWAtOltk1ucZYbK4e+rgQFfhgucTvl8Z4tR46NqA0vgASzTUht9Cbg1UI75xv0ITwpaOaQjmPQA1HjCJMskF2fcPZib2UiGEovXHplGIQ9JSPC+YacZenbinmbstd+6AFTDGjZMbZyonEWzPSm4TNEIPTkLFyQ6aIQoAQq0XciUEFoGioYPsIe2OwMDhjIws9tQE+dn2SnsC/Ztbkf4moi3j/AICv7B7zVbvv98U6/vyXEc8Vx3cbkhmnSUDHndfJGtLGLZAg2/e9VhUnIglK0Vn3PuW9SXNH/9VdrexNLWBLqX0U57r7eZd+e0OtCOtJXyJyC3io6ih8V6aLx+3Li1PeMoiHOuLBUH8AcKkNOVQAABtqSURBVDBfw9wY7JumAUvWzMNHi+4MxND3NoaCY1zxYjjj+TDjKq6I1rRkKPtwtgzESw6fADhuANhUg3PdaNExbBiAPQS404Ht8+46Ag4syo4StLIuAM9+o+Gt/+XLn/7F8/t4fadB4D8H8KMAfgLqP/gffbsvQEQ/RUS/QES/kNW57Hd91XffBUQwvCwYXwrCeSMSSUAfA3K2yYCblNgicgablgaAzrr9e6vhBNy2y3kD/bm7sVwPHMIdZaebSW27cw2dw7/UqLdL7FmBbtat877WgLs89aDzap3wcj3gvgx42JUVFdov2B+avDLmOeF2nbDYtMKDymCb35WTzjWZgOgmCOKTCA9ua9P3uSc/iX8e9v72nxtgCZgFAYhSiEnQ/y7aA6ANDzAH9ZNgoI4CSYBEIJ4bym+9/Ymvlc/S+o5gwyLyjt8mov8SwP9k//0agB/cPfQP2H0f9xrfc0PSb2XF9+7Bz4+YPoxYnwfUI7QRGBRvShXgQpCsp4t+T58rQrb51ZLc5+DVcPK1MU5l2DrzNkdHqBdoOKfRNtH5eUYwWe5qRiVb7PaNnncnqQeFKgreYWyNxCaEFCrOJWHgapTebWIwcOmjxrP9VaiqC/P9MqIJcKbUS4wLsVRDB+67/gAu+g8ehGakfrs2Vl9IAxntXYb8c9J037QadplXS4rRwO5+KUou0tGvaYkEwc2XgcPXH564Q6+t7ygTIKIv7f77JwH45OCvA/hTRDQS0Q9DvQh//pO9xe/tqr/0qwjfeInp3QXT+5YN+CnD6I1Ax9ZIFMhghBY72Tio1DizcgLWEgwL0HDKylxh4wvkqs2xItsp6Es3dMBSt867p+Ep1L7ZcuMuC+7Mws4fMPHO2pTmK0IYWIPOXGMvLx7y2KXLPT9pbRuRAsBpSWquCs1QzkXJRvd5xFxTnxbcl3GXuWhJ48pCsjv91Z6MurNQtd+jOO4fuvG7mIhsBiNo9qcwWHP/G1XHElgJkASSBC0Cb/7iGfJ3fvG7cdn8vl7fqRfhHyWin4BWXF8G8K8CgIj8IhH9VQD/D9Sy/M+KyO87bFb5ja9gqA3Hqy9hfTbg4QAFA0UAoheU71UJsrHXPHU12HAIrTsQlxrUVNM2+hCq1r1tsyfz3sC5JOUOkGAMtTfVfO278LVuG4dIkXy+CbuZiZUEPvv3bnzLQzftCPY+XTAU0CwDDMjYwFG1E4iANu2DFUNEOsOwWLABVLC0CfVeAgCdMNjvuUcJwkRFlxKwrpungysL96AAbQAKo4O2AFhZZpgGL9VslAsA4USg+pQDfNz6VL0I7fF/DsCf+yRv6vfEShFxrrj+WkW+CljeFDMqUdhwnxIQNAisjFoT2lgRx4JaCbVGrKRp6jBUHNKMc064g08FtpZCto1+SPnibXQQzaoZhG+0fbpdDZkYDZG4lNhViZ2h6Cftvh9B3NAs9S6NcQpDHx+OoWBKBfeAsiiD4Pq4IIaGKW5w4e6gZF6KDEUa5hq6eWoV7nyByHqEl8ZYrNcRQ0MuAUuOyKu6OsekdM7Wdp6CAhBx5wR0LEAhgHV060EAQHeXokL4+/6rd1B/+feMQfbvqfVoqcS/0yq//hsYzjPwh74fb/zSgPf/cMDyhkBiA6+kDjYFgDCaZQO0mACmEVt0k+mFPNeAW1YxkZevjuDQMI4Zo2kP1BYxxNq19+/X0Tr0Skf2BhlDN/WrZepptbISgaUEnHLCG9MZAXriQ9ADwDnr2O9mXHBIGQ/rgFIZgQXX44Ip5E1PoAW03rXU8uYwaIBy5qOe9EpvzjVcBKelBrz9cNNHiTZg6c3AXAPu51EDh2VBzhaEAJUCKgA2nICzCbt3gDE30bRpSwVmTiid9CQON35av+16CgK/zSrvfAPpdAb94R/Gzc0Ryxvo4BkA26lj+oRU7V8GYqrKgLM0tlXC6TRinMx/oCo1OefYQTFMKggy9pHgpQnnUgOa4wZsMQlqqAis6kZ3bUIy/P6F8rDV3kNsHQx0Xl3gRN+TThZUNOTVMmHJyWjVZOPOiGivOxcVQTmYlHqxIOMuzbmq2UcKFSnqe05mqba2gNOq4CfmhlL1vTd7rlRGWYKe8ibfzhbomBtaIkjZbX5AP3cBUGizmof+vf7+P/9llLd7L/tpvbYePXfgt10ikKoX4fVXVwy3m621wERuClTmilR8A1DxihgrhkF1BFT3TtAaYV1DZ7z1i14I83nAwzxgtgbfIWaM+9SdxObnmyjpPhh4bZ9zwP0y4GHdpHMcjtv1DWvAaRk6j6E2VT92gVOHIxOJei8Efb+lbhMIR/tlG1Ouxv+f14TzMmBZEkTQ63/6/9s721hLr6qO/9bez8s596XTznQY2iltSq20gFrBACZKfCEKhKQaI5YvEmOCJPDRD5j4wfhFPkiMBjGBWAGNIhorBKuAGBUl1Za09A36Bq1t03b6QufOnXvOeZ699/LD2s9zzgwdmOm9M2fm3uf/5d7z3Le97znPOnut9V//P52D8LwW4ETN1Tl1vP9uzNf+p9LafEDIRrApZTagUyjmXQHJGhAScz0gWhfHTx3X//HTFgB0OBKcCkMQ+AHQ2YzyO8/gJ4F9jySKY4JrF3rUDvxMkODA2wsznSSAoSp4n9AkhGlpHn8ZMZpBSYxmZa65xdcm3zvtzG256I1LfD5Cd+29wqfeIXnalDTBfw8hqV8PdjLolHzDwve10bPZ1HOxk8zb79OR3JnoCD6zUDAL2W8gOdrWCntg795gYsBR56rBMVf5u8/NHcjN79PeH8CCQZr53mYM6GcJ5mQtsRStzVLifXHQ0rohAHx/DOnAD4CGQHj6GcqqZK0+yOyiEdODmYDSEYii4KZCdAKVdQaaxvctwu44D6CtvZh9Hecngjx2bPLjjsLN3ykXBUi8KC1WxCsXevFd2630kejNmWcmNsk4Km22304NHYPR9iaiJ6Qs3RSgCX/k/DtKHwS6QNBxHkKcdy3abAnW7WWBGjAPNgv0aN+RhHQ+MtwbhfRWYQJRIThSJwef3YRsIYpG7W94ifPOgWuFg3fFIQCcBoYgcDpQJTz2OO7wfi56PKBFwewSKwD2gSCYfVmsk1mYt56g0rMH+xl3UWjF5MN9lxYYwUWdHdVHZeiLglVuFaaFG72N80r84ihuVXQFtIq2KaxN6fSEtCF0NOQ8rGT+fva1rqXoXepFVE3bX3E+y6IxJyh1H6NazaBtzGehy9914V1fRIlu3ucvsuJypyl4wq2aWYGS6xE0oCKoN11B7ZygKhvW6hjDLtuRuRbWnlBWbvmfs/SC2F0YgsAZQEvH6NkpMGIjFkwPCIyNP6DO1Ipj4/CriTYU6FaBesWv5qp6tjrDKyQxHz01xRyCvci7Q3tPMc68/FksmOWBGrDaQK9BoALJhEkLl2iCZzaztKMNNg3YEXW8S6zUDdO2YDarbMip6MhHeQKwEppQ9DULMudhcbApJjvBhGwFHoOf27Q5S01SkjxSLVlf0ViBHeyERF8zUMkaAEnzbED3fUIsMhFIOoFWcqrCfEJQ7WfGzykX/+X5P7N2vmCoCZwB3H/ciTph9MyEi/4vMHreNAb81F6bsbIOQYrOZtq9Io0jToo8467Gec/jsKnx6FYBM9d3HNpgVftZW9pxPCMkx8a0ZhYKVssGEWWac/FREVitmj7HHlVtJtpI37NvW5NJ747wKbm+NuFc6iXPgtop49DasV4UhXzDd3l8G62n3zR+ns9jXRFZeEWlbBUe89/s9BK8zPkLxgFYOAcoWSsgd1vUFI8RTNmp6ZSH6YuI6ub1GWRO5BpwehhOAmcI+do38K97DaNnBUk1LhRM9wspQnORaeGH42U/gqtekaknFAlfJWLj+m6BCnZ8nTk0KLISzNOwuyGT66v8/Y2bks3ytzVHJyNCcqxULfvHWzhRjhxbw4myb21iR/pMwS0Kq1W0ref4rKIuIiujhibbpx+d2HDRxaMJG7MR+0dblEVkGqV3XeoGe6ZtQdta/u+cDUw51azvZzp/bQu6oBEguQsxaQtzZ/KxX1tfb9BsDJvJPqkyRSfzgwRE7JRAMqJWzww0FoIkuOSByNrfDWnAmWAIAi8D8f4H8ddfSzEuWH8cii1LDWIlpDHWW4/zo786ha2COIrIKM5FMoMgrTM9gq7dldtmqnZ8TkofCKrChn2ObK0zLlvqMqDNfBDHoVRFZBY8JTCuWiZNaT34lJV8HUynJamy4aJRJgCF6NmY1r2oyYuzseXrESQ4QuMJmVswnZZ9Xp562iS9pFrvAgw5VXC47EbshBPMVp1LpOTno9T90JL08xmiNh5sYiHZYhzN/7/cFQhw2X/PKP/trrP2vO9WDEHg5UCV9NCjVHoV4dUXs3KkpdzyIJ5UO+I4zamrBb3RKTPfW2V1CroS8/3Sva6jpy5DX0VvQkEI1iKr6sC+lQkhOV6cjAEoszPQVlv1N3CRmYn+JEEQ7xOjKubf6Qg4qLC/l5TJrESw4DGLnq1jNZKk5+g3wdKKOC1spl/zO7MK4k373zadNRKi5fGaC4Vg7cI2egTfqwhJZiR2hcQuMPYWY1HBc+LYcDDpdz8VXAtXfWETd/dDpHTBjaosHUMQeJnQtkEfe5KVmJi8ej+j5+0dNXnP9BVikmRqlew0SlAnaAWaLIOV8lHWkcdoIQVH0xaMq7av0LdNYT3yfHSeZAqxZrpwV/mf5iEi70yEY9KUxMz7dU6JcS5cYsdve5eGTvI8nTCl1wQT8fRTIY6tqDmdlbTTwvZQkYk70Bl8aHL9LERHEQZOSG9in57Mb1YR+tHrfggrkqXdcq6/UPwza3F795cIV916HHfft0nT6Vl5rnc7hiCwDaStLdwTTzH2jsmV+6iOBtaeFEQdzb5OiUhN3EIcdGaZ0aTKe3prnonvWYVZobdtvRGI2szUS8KsLXp/w6Qn8geMHKTU5Vy1B05slbfRagBdamC/R0hqJ4Xu5m2jOS0VE8FFaClpZi6LeQgq2tt7pWzmkeKcUt25/0g2VQnBoWrUYHGJsoh9UBBZEFVZEG/t+v6530hvPy4gmRx0+Vdb3F0PDgFgGxiCwDaRtraQhx+j2ncdbrMBt4ZvPZuXeSavVGJt+a3bFOIqdiJo5uQWNKcMZaKoTINga5opv12OnE8O1LG/YWJyyEIZvDtad+YedRmYtgVNnsqzd9pkrkittxuxjD1tuFPyAetQxGj6fH5q8uso1tL0oJWdJMRnfUSX8/ToLNXpyD5i3YKkQmqLbA4quGJ+SujIQouQbtpI8/hvp+PQ/VK1duz++5Tqy3cOKcA2MQSBHYC2Ddx2NwmoRq+nfFFwYUyqCuLICEV+Ikj0hEsUioQWRjV2TRbJWUuMx42N1G5VuEIpyoArHHFiBTBxytpoRhvnsmSdAaiq9BwBVaH0gZiEViyd6IaW+9Hi1lMUKY8f0/P3Y3IEtQEn19UsOrpAyv4KXvtZAJ8HkhJdsQ+7i3OQQIWUJwO7YqFG6SXZndN8SlgoDHZBYGFcWhZahy7A6uPCvr8auAA7gSEI7DD0jntxP3Y9oyMzLgZcUzB5hQUC14I77knjNJ9ElHl/e/5LrOAW2sJaioAWc3vwSWPTh1IGREJ/hJec24so09ZOAGVpBKDOlOOEnjymRdBGT9N4NLleDCUET3Xc6gGz/Yk0SvgtS2HIBqDO24y/E+2FVZxLSGHU3qTz9l3vAWjRI6/JiEadg3BXo7DBoI5hyZwIhJ2e6heEV3xs8A3YKQxkobOA9I1v4jdmjJ5tOHj3hEvvTpTHhDhSXCMUGx43tXfVOE7oOOK8MpnaFF+92lilfKuAiTfSjDe77Y2J6Qh0rb0m+H6oCOx+KX1iXAaj9mZps7oKrNYNddnmanw2PI3G+deODJSEumyJGxX1dzEyTme7FgU/ld5/sfMEcD7NC3pOWRk1VHXoOf6uMGUi816zYBWjiZd2w0Y9xP5el+lIsgEt3wguwL6H4fI/HALATmI4CZwlpHu/hX/9dUwPr7H+8Cbj52peuK5muj/7GFSddZkiVbIWWRKaWdb7T5lHkFt0UtlN2zQFMU8MdupEpY+9AahZlc1dhkWMO7BaNxQusTm1CcGyCozrxkQ91YKHc4mqCuaL+KwnjCGs21yEm1rqItFaeCk4GspeS7Guc3ckOaZNSYyCLyLeGzU46IkvtRjN8h0HbqWxKcu26AuPqVRcEJMFy4Khh24P1LfecS6fxj2BIQicRaT7H6J+wCNVSXntVRy8M3D88JjNyz2zA9Cu2QSiNg4dQT1q2dqsUcQKbLWgM2fvxkFIpcORmLUlxVqkLgNtdGxOzc688Kb+0+YZ/a71V/i5WUlH0OlUerrkQADvbepw4/gIPxVmlyjtxWbS6je9qSkBfssRXEEYJVIljMZN37GImixQtZ0Ec2YUugS1iatophPTOCQJLRXFyFqfUqZ8Hfu7rQmHXP7lDbjnIXSYCtxxnI7Q6M3Au4Ajqvr6fO1vgdfkb7kYeFFVb8hORd8EHshfu01V37/Ti75gkCKaIhpa3COPoz96DStPTfFNzWbj2TrkaNVYge3RmjAu0NaZnVlhN496Y8bROrSOqAiptaN0XUSqImYWnvQ3YhM8pc9tuGgjvxNK6iLgRQmtp6jspgttQTMpQYWqzvyEzRp3IJFWktl+RSHVFjhQy9nLo5503JMOJerSTg+z0HkFGMNPg81EODWpNecjSLYWb10vCKqtEL3tu7N5N5YQvOqWp5FpQ3r+BdLs/PCn2G04nZPAJ4GPAp/uLqjqr3Wfi8hHgKML3/+Iqt6wUwvcFVAlHTtG+e2nmfzIFVTfbVgPZS4aCu1FjljbzY1TdJRMO69rDWaYWMlcgyAmwTvpq/tgnIFupBjy5F4WHA3J9dOJRZHVhZvCdPrLxPo432RTR1qLHP6So8jHcQkB32blJCe4JvHou8b4OgeT6Aj5d8fgKOpA0NKGpHyiqBNFkWg7slTXLUggTVYULhfe5Z1y5T9PiY88BkML8KxiW16EIiLAu4Gf29ll7U6Ep56meuUBppetUGxF1p9UyknB5FLHbJ8Q1pX2omguRmK22pJvGC2sTqC5wm41AMdW42jbLElWtayUDbNYsJFzf2MQzk8IbesRp5TFgkGIU8o6sK+e8sJkhUNfc4SxsP7Fe0nHjp1yPwcPv4XjT63RFmsms7YG0+snaBLGazNS9MSJFR1T4Y3ukMRchcPCbEC0a+URx/pj1gJ0reL+a5gDOBfYbk3gp4FnVPWhhWtXi8idwAbwu6r61W3+jV0FvfM+0pVvws0i1bPHKTdqfDPGTx3TKITVOeuOJJltiB3LBWiyjBnQtAWh9cTGg1NmVUu1ai5FnRmJSOgJRG30RsDJgaH0iS1vnIV2WvDgQ5fjj3muyf339FIbWMDaZ29jbeGx/+FreOxXDqEOmvUaPTztLcTjBFKRJcL6QqMwPpLbgg4O3N9S/9PtO/jfHnA62G4QeA/wNwuPnwKuVNXnReSNwD+KyOtUdePkHxSR9wHvAxixss1lXFgYf85MmSJQXHGYcu2VVEdh9N0CiZ6ty60jYJp5Yg5HTm2CdmoDOZGS6LLHVp7oO15VPF+vmB7BzAw8CpdQVSazirb1vcLvtCmZCbSTEjYK9t/jOPCJ7ZFv4oOPcMUfPNI//s6HfxKfyPRo3/MhXDDGn5/C4Y/f831PGwPOPuR0qq05HfhCVxjM1wrMZ/CNqvrEKX7u34HfVtXv29e5SPbrm+XnT3/VuxBSFHDDdSDC9NIRcexoVh3T/cLkkNIeCEidkBdL3EyIa/l9OkhfR3CjrAw08+AVX2VN/+T6ekM9bolRaI9XIPBDfxFxX71ziTsfcK7wr/r3X1fVnzj5+nZOAm8DvrUYAETkIPCCqkYReTXmRTjYvpwGNAT4+n1s/uqb8U1idGTGShNJlafdVzK9xBPqwkZ69wlh05sLeguphNkBQatok3+bjriS0DLZANLxAjd1pNVIrAJX33T3src74DzCy/IiVNU/B27ixFQA4K3A74tIi6WU71fVF3Z2ybsYqqz/gx2aZm/7cVzr8VuB+vkZ1QtiVluFEGtPu+qJlRBroVkTXOuIR40v4GaCbjiu/uiL6KM5RqcEzoHID8z1B+wtnFY6cLYxpAPfC7e+bjeuKr0ccDf7K4J4bxx8ceCdpROdckd+TuNzz9sJY8AAzk46MOAsYiiWDThXGAaIBgzY4xiCwIABexxDEBgwYI9jCAIDBuxxDEFgwIA9jiEIDBiwxzEEgQED9jiGIDBgwB7HEAQGDNjjGILAgAF7HEMQGDBgj2MIAgMG7HEMQWDAgD2O82KUWESeBY4Dzy17LWcZl7K797jb9wcX9h6vUtWDJ188L4IAgIjc8VKzzrsJu32Pu31/sDv3OKQDAwbscQxBYMCAPY7zKQh8fNkLOAfY7Xvc7fuDXbjH86YmMGDAgOXgfDoJDBgwYAlYehAQkbeLyAMi8rCIfGjZ69kpiMijInKPiNwlInfka/tF5Msi8lD+eMmy13kmEJGbReSIiNy7cO0l9ySGP8nP690i8oblrfz0cIr9/Z6IPJmfx7tE5J0LX/udvL8HROQXl7Pq7WOpQUBEPPCnwDuA1wLvEZHXLnNNO4yfVdUbFlpKHwK+oqrXAl/Jjy8kfBJ4+0nXTrWnd2DmM9didnN/do7WuB18ku/dH8Af5efxBlW9FSC/Tm8CXpd/5mP59XzBYdkngTcBD6vqt1W1AT4D3LjkNZ1N3Ah8Kn/+KeCXlriWM4aq/idwspnMqfZ0I/BpNdwGXCwil52blb48nGJ/p8KNwGdUdaaq3wEexl7PFxyWHQQOA48vPH4iX9sNUOBLIvL1bL4KcEhVn8qfPw0cWs7SdhSn2tNuem4/mFOamxdSuF2zv2UHgd2Mn1LVN2DH4g+IyFsXv6jWltlVrZnduCcsjbkGuAFz3f7Icpez81h2EHgSeNXC4yvytQseqvpk/ngEuAU7Kj7THYnzxyPLW+GO4VR72hXPrao+o6pRVRPwCeZH/l2xP1h+ELgduFZErhaRCiu0fH7Ja9o2RGRVRNa7z4FfAO7F9vbe/G3vBT63nBXuKE61p88Dv567BG8Bji6kDRcMTqpj/DL2PILt7yYRqUXkaqwA+r/nen07gaV6EapqEJEPAl8EPHCzqt63zDXtEA4Bt4gZhBbAX6vqv4jI7cBnReQ3gceAdy9xjWeMl3KoBj7MS+/pVuCdWMFsC/iNc77gM8Qp9vczInIDluY8CvwWgKreJyKfBe4HAvABVY3LWPd2MTAGBwzY41h2OjBgwIAlYwgCAwbscQxBYMCAPY4hCAwYsMcxBIEBA/Y4hiAwYMAexxAEBgzY4xiCwIABexz/D0X3tmcMq00GAAAAAElFTkSuQmCC\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "M0i1H3FswM8W",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "b3980441-39bc-4e98-87d6-298ae7c90de5"
      },
      "source": [
        "X_train.shape,Y_train.shape,X_val.shape,Y_val.shape"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "((540, 192, 192, 4),\n",
              " (540, 192, 192, 4),\n",
              " (180, 192, 192, 4),\n",
              " (180, 192, 192, 4))"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 20
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "tvcU4hSRFuJM"
      },
      "source": [
        "## 2D-VNet model\n"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "FZ3asv7Pqsjr"
      },
      "source": [
        "\n",
        "def dice_coef(y_true, y_pred, epsilon=1e-6):\n",
        "    intersection = K.sum(K.abs(y_true * y_pred), axis=-1)\n",
        "    return (2. * intersection) / (K.sum(K.square(y_true),axis=-1) + K.sum(K.square(y_pred),axis=-1) + epsilon)\n",
        "\n",
        "def dice_coef_loss(y_true, y_pred):\n",
        "    return 1-dice_coef(y_true, y_pred)\n",
        "  "
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "MjcJBD44cMh8",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "914e029b-8377-4d25-9b40-97d751d3affa"
      },
      "source": [
        "def resBlock(conv,stage,keep_prob,stage_num=5):\n",
        "    inputs=conv\n",
        "    \n",
        "    for _ in range(3 if stage>3 else stage):\n",
        "        conv=PReLU()(BatchNormalization()(Conv2D(16*(2**(stage-1)), 5, activation = None, padding = 'same', kernel_initializer = 'he_normal')(conv)))\n",
        "    conv_add=PReLU()(add([inputs,conv]))\n",
        "    conv_drop=Dropout(keep_prob)(conv_add)\n",
        "    \n",
        "    if stage<stage_num:\n",
        "        conv_downsample=PReLU()(BatchNormalization()(Conv2D(16*(2**stage), 2, strides=(2, 2),activation = None, padding = 'same', kernel_initializer = 'he_normal')(conv_drop)))\n",
        "        return conv_downsample,conv_add\n",
        "    else:\n",
        "        return conv_add,conv_add\n",
        "        \n",
        "def up_resBlock(forward_conv,input_conv,stage):\n",
        "    conv=concatenate([forward_conv,input_conv],axis = -1)\n",
        "  \n",
        "    for _ in range(3 if stage>3 else stage):\n",
        "        conv=PReLU()(BatchNormalization()(Conv2D(16*(2**(stage-1)), 5, activation = None, padding = 'same', kernel_initializer = 'he_normal')(conv)))\n",
        "    conv_add=PReLU()(add([input_conv,conv]))\n",
        "    if stage>1:\n",
        "        conv_upsample=PReLU()(BatchNormalization()(Conv2DTranspose(16*(2**(stage-2)),2,strides=(2, 2),padding='valid',activation = None,kernel_initializer = 'he_normal')(conv_add)))\n",
        "        return conv_upsample\n",
        "    else:\n",
        "        return conv_add\n",
        "\n",
        "def vnet(pretrained_weights = None,input_size = (192,192,4),num_class=1,is_training=True,stage_num=5,thresh=0.5):\n",
        "    keep_prob = 0.1\n",
        "    features=[]\n",
        "    input_model = Input(input_size)\n",
        "    x=PReLU()(BatchNormalization()(Conv2D(16, 5, activation = None, padding = 'same', kernel_initializer = 'he_normal')(input_model)))\n",
        "    \n",
        "    for s in range(1,stage_num+1):\n",
        "        x,feature=resBlock(x,s,keep_prob,stage_num)\n",
        "        features.append(feature)\n",
        "        \n",
        "    conv_up=PReLU()(BatchNormalization()(Conv2DTranspose(16*(2**(s-2)),2,strides=(2, 2),padding='valid',activation = None,kernel_initializer = 'he_normal')(x)))\n",
        "    \n",
        "    for d in range(stage_num-1,0,-1):\n",
        "        conv_up=up_resBlock(features[d-1],conv_up,d)\n",
        "    if num_class>1:\n",
        "        conv_out=Conv2D(num_class, 1, activation = 'softmax', padding = 'same', kernel_initializer = 'he_normal')(conv_up)\n",
        "    else:\n",
        "        conv_out=Conv2D(num_class, 1, activation = 'sigmoid', padding = 'same', kernel_initializer = 'he_normal')(conv_up)\n",
        "    \n",
        "    \n",
        "    \n",
        "    \n",
        "    model=Model(inputs=input_model,outputs=conv_out)    \n",
        "\n",
        "    model.compile(optimizer = Nadam(lr = 2e-4),loss=dice_coef_loss,metrics=[dice_coef,'accuracy'])\n",
        "    \n",
        "    if(pretrained_weights):\n",
        "    \tmodel.load_weights(pretrained_weights)\n",
        "    return model\n",
        "model=vnet(input_size = (192,192,4),num_class=4,stage_num=5)\n",
        "model.summary()\n"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Model: \"model\"\n",
            "__________________________________________________________________________________________________\n",
            "Layer (type)                    Output Shape         Param #     Connected to                     \n",
            "==================================================================================================\n",
            "input_1 (InputLayer)            [(None, 192, 192, 4) 0                                            \n",
            "__________________________________________________________________________________________________\n",
            "conv2d (Conv2D)                 (None, 192, 192, 16) 1616        input_1[0][0]                    \n",
            "__________________________________________________________________________________________________\n",
            "batch_normalization (BatchNorma (None, 192, 192, 16) 64          conv2d[0][0]                     \n",
            "__________________________________________________________________________________________________\n",
            "p_re_lu (PReLU)                 (None, 192, 192, 16) 589824      batch_normalization[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "conv2d_1 (Conv2D)               (None, 192, 192, 16) 6416        p_re_lu[0][0]                    \n",
            "__________________________________________________________________________________________________\n",
            "batch_normalization_1 (BatchNor (None, 192, 192, 16) 64          conv2d_1[0][0]                   \n",
            "__________________________________________________________________________________________________\n",
            "p_re_lu_1 (PReLU)               (None, 192, 192, 16) 589824      batch_normalization_1[0][0]      \n",
            "__________________________________________________________________________________________________\n",
            "add (Add)                       (None, 192, 192, 16) 0           p_re_lu[0][0]                    \n",
            "                                                                 p_re_lu_1[0][0]                  \n",
            "__________________________________________________________________________________________________\n",
            "p_re_lu_2 (PReLU)               (None, 192, 192, 16) 589824      add[0][0]                        \n",
            "__________________________________________________________________________________________________\n",
            "dropout (Dropout)               (None, 192, 192, 16) 0           p_re_lu_2[0][0]                  \n",
            "__________________________________________________________________________________________________\n",
            "conv2d_2 (Conv2D)               (None, 96, 96, 32)   2080        dropout[0][0]                    \n",
            "__________________________________________________________________________________________________\n",
            "batch_normalization_2 (BatchNor (None, 96, 96, 32)   128         conv2d_2[0][0]                   \n",
            "__________________________________________________________________________________________________\n",
            "p_re_lu_3 (PReLU)               (None, 96, 96, 32)   294912      batch_normalization_2[0][0]      \n",
            "__________________________________________________________________________________________________\n",
            "conv2d_3 (Conv2D)               (None, 96, 96, 32)   25632       p_re_lu_3[0][0]                  \n",
            "__________________________________________________________________________________________________\n",
            "batch_normalization_3 (BatchNor (None, 96, 96, 32)   128         conv2d_3[0][0]                   \n",
            "__________________________________________________________________________________________________\n",
            "p_re_lu_4 (PReLU)               (None, 96, 96, 32)   294912      batch_normalization_3[0][0]      \n",
            "__________________________________________________________________________________________________\n",
            "conv2d_4 (Conv2D)               (None, 96, 96, 32)   25632       p_re_lu_4[0][0]                  \n",
            "__________________________________________________________________________________________________\n",
            "batch_normalization_4 (BatchNor (None, 96, 96, 32)   128         conv2d_4[0][0]                   \n",
            "__________________________________________________________________________________________________\n",
            "p_re_lu_5 (PReLU)               (None, 96, 96, 32)   294912      batch_normalization_4[0][0]      \n",
            "__________________________________________________________________________________________________\n",
            "add_1 (Add)                     (None, 96, 96, 32)   0           p_re_lu_3[0][0]                  \n",
            "                                                                 p_re_lu_5[0][0]                  \n",
            "__________________________________________________________________________________________________\n",
            "p_re_lu_6 (PReLU)               (None, 96, 96, 32)   294912      add_1[0][0]                      \n",
            "__________________________________________________________________________________________________\n",
            "dropout_1 (Dropout)             (None, 96, 96, 32)   0           p_re_lu_6[0][0]                  \n",
            "__________________________________________________________________________________________________\n",
            "conv2d_5 (Conv2D)               (None, 48, 48, 64)   8256        dropout_1[0][0]                  \n",
            "__________________________________________________________________________________________________\n",
            "batch_normalization_5 (BatchNor (None, 48, 48, 64)   256         conv2d_5[0][0]                   \n",
            "__________________________________________________________________________________________________\n",
            "p_re_lu_7 (PReLU)               (None, 48, 48, 64)   147456      batch_normalization_5[0][0]      \n",
            "__________________________________________________________________________________________________\n",
            "conv2d_6 (Conv2D)               (None, 48, 48, 64)   102464      p_re_lu_7[0][0]                  \n",
            "__________________________________________________________________________________________________\n",
            "batch_normalization_6 (BatchNor (None, 48, 48, 64)   256         conv2d_6[0][0]                   \n",
            "__________________________________________________________________________________________________\n",
            "p_re_lu_8 (PReLU)               (None, 48, 48, 64)   147456      batch_normalization_6[0][0]      \n",
            "__________________________________________________________________________________________________\n",
            "conv2d_7 (Conv2D)               (None, 48, 48, 64)   102464      p_re_lu_8[0][0]                  \n",
            "__________________________________________________________________________________________________\n",
            "batch_normalization_7 (BatchNor (None, 48, 48, 64)   256         conv2d_7[0][0]                   \n",
            "__________________________________________________________________________________________________\n",
            "p_re_lu_9 (PReLU)               (None, 48, 48, 64)   147456      batch_normalization_7[0][0]      \n",
            "__________________________________________________________________________________________________\n",
            "conv2d_8 (Conv2D)               (None, 48, 48, 64)   102464      p_re_lu_9[0][0]                  \n",
            "__________________________________________________________________________________________________\n",
            "batch_normalization_8 (BatchNor (None, 48, 48, 64)   256         conv2d_8[0][0]                   \n",
            "__________________________________________________________________________________________________\n",
            "p_re_lu_10 (PReLU)              (None, 48, 48, 64)   147456      batch_normalization_8[0][0]      \n",
            "__________________________________________________________________________________________________\n",
            "add_2 (Add)                     (None, 48, 48, 64)   0           p_re_lu_7[0][0]                  \n",
            "                                                                 p_re_lu_10[0][0]                 \n",
            "__________________________________________________________________________________________________\n",
            "p_re_lu_11 (PReLU)              (None, 48, 48, 64)   147456      add_2[0][0]                      \n",
            "__________________________________________________________________________________________________\n",
            "dropout_2 (Dropout)             (None, 48, 48, 64)   0           p_re_lu_11[0][0]                 \n",
            "__________________________________________________________________________________________________\n",
            "conv2d_9 (Conv2D)               (None, 24, 24, 128)  32896       dropout_2[0][0]                  \n",
            "__________________________________________________________________________________________________\n",
            "batch_normalization_9 (BatchNor (None, 24, 24, 128)  512         conv2d_9[0][0]                   \n",
            "__________________________________________________________________________________________________\n",
            "p_re_lu_12 (PReLU)              (None, 24, 24, 128)  73728       batch_normalization_9[0][0]      \n",
            "__________________________________________________________________________________________________\n",
            "conv2d_10 (Conv2D)              (None, 24, 24, 128)  409728      p_re_lu_12[0][0]                 \n",
            "__________________________________________________________________________________________________\n",
            "batch_normalization_10 (BatchNo (None, 24, 24, 128)  512         conv2d_10[0][0]                  \n",
            "__________________________________________________________________________________________________\n",
            "p_re_lu_13 (PReLU)              (None, 24, 24, 128)  73728       batch_normalization_10[0][0]     \n",
            "__________________________________________________________________________________________________\n",
            "conv2d_11 (Conv2D)              (None, 24, 24, 128)  409728      p_re_lu_13[0][0]                 \n",
            "__________________________________________________________________________________________________\n",
            "batch_normalization_11 (BatchNo (None, 24, 24, 128)  512         conv2d_11[0][0]                  \n",
            "__________________________________________________________________________________________________\n",
            "p_re_lu_14 (PReLU)              (None, 24, 24, 128)  73728       batch_normalization_11[0][0]     \n",
            "__________________________________________________________________________________________________\n",
            "conv2d_12 (Conv2D)              (None, 24, 24, 128)  409728      p_re_lu_14[0][0]                 \n",
            "__________________________________________________________________________________________________\n",
            "batch_normalization_12 (BatchNo (None, 24, 24, 128)  512         conv2d_12[0][0]                  \n",
            "__________________________________________________________________________________________________\n",
            "p_re_lu_15 (PReLU)              (None, 24, 24, 128)  73728       batch_normalization_12[0][0]     \n",
            "__________________________________________________________________________________________________\n",
            "add_3 (Add)                     (None, 24, 24, 128)  0           p_re_lu_12[0][0]                 \n",
            "                                                                 p_re_lu_15[0][0]                 \n",
            "__________________________________________________________________________________________________\n",
            "p_re_lu_16 (PReLU)              (None, 24, 24, 128)  73728       add_3[0][0]                      \n",
            "__________________________________________________________________________________________________\n",
            "dropout_3 (Dropout)             (None, 24, 24, 128)  0           p_re_lu_16[0][0]                 \n",
            "__________________________________________________________________________________________________\n",
            "conv2d_13 (Conv2D)              (None, 12, 12, 256)  131328      dropout_3[0][0]                  \n",
            "__________________________________________________________________________________________________\n",
            "batch_normalization_13 (BatchNo (None, 12, 12, 256)  1024        conv2d_13[0][0]                  \n",
            "__________________________________________________________________________________________________\n",
            "p_re_lu_17 (PReLU)              (None, 12, 12, 256)  36864       batch_normalization_13[0][0]     \n",
            "__________________________________________________________________________________________________\n",
            "conv2d_14 (Conv2D)              (None, 12, 12, 256)  1638656     p_re_lu_17[0][0]                 \n",
            "__________________________________________________________________________________________________\n",
            "batch_normalization_14 (BatchNo (None, 12, 12, 256)  1024        conv2d_14[0][0]                  \n",
            "__________________________________________________________________________________________________\n",
            "p_re_lu_18 (PReLU)              (None, 12, 12, 256)  36864       batch_normalization_14[0][0]     \n",
            "__________________________________________________________________________________________________\n",
            "conv2d_15 (Conv2D)              (None, 12, 12, 256)  1638656     p_re_lu_18[0][0]                 \n",
            "__________________________________________________________________________________________________\n",
            "batch_normalization_15 (BatchNo (None, 12, 12, 256)  1024        conv2d_15[0][0]                  \n",
            "__________________________________________________________________________________________________\n",
            "p_re_lu_19 (PReLU)              (None, 12, 12, 256)  36864       batch_normalization_15[0][0]     \n",
            "__________________________________________________________________________________________________\n",
            "conv2d_16 (Conv2D)              (None, 12, 12, 256)  1638656     p_re_lu_19[0][0]                 \n",
            "__________________________________________________________________________________________________\n",
            "batch_normalization_16 (BatchNo (None, 12, 12, 256)  1024        conv2d_16[0][0]                  \n",
            "__________________________________________________________________________________________________\n",
            "p_re_lu_20 (PReLU)              (None, 12, 12, 256)  36864       batch_normalization_16[0][0]     \n",
            "__________________________________________________________________________________________________\n",
            "add_4 (Add)                     (None, 12, 12, 256)  0           p_re_lu_17[0][0]                 \n",
            "                                                                 p_re_lu_20[0][0]                 \n",
            "__________________________________________________________________________________________________\n",
            "p_re_lu_21 (PReLU)              (None, 12, 12, 256)  36864       add_4[0][0]                      \n",
            "__________________________________________________________________________________________________\n",
            "conv2d_transpose (Conv2DTranspo (None, 24, 24, 128)  131200      p_re_lu_21[0][0]                 \n",
            "__________________________________________________________________________________________________\n",
            "batch_normalization_17 (BatchNo (None, 24, 24, 128)  512         conv2d_transpose[0][0]           \n",
            "__________________________________________________________________________________________________\n",
            "p_re_lu_22 (PReLU)              (None, 24, 24, 128)  73728       batch_normalization_17[0][0]     \n",
            "__________________________________________________________________________________________________\n",
            "concatenate (Concatenate)       (None, 24, 24, 256)  0           p_re_lu_16[0][0]                 \n",
            "                                                                 p_re_lu_22[0][0]                 \n",
            "__________________________________________________________________________________________________\n",
            "conv2d_17 (Conv2D)              (None, 24, 24, 128)  819328      concatenate[0][0]                \n",
            "__________________________________________________________________________________________________\n",
            "batch_normalization_18 (BatchNo (None, 24, 24, 128)  512         conv2d_17[0][0]                  \n",
            "__________________________________________________________________________________________________\n",
            "p_re_lu_23 (PReLU)              (None, 24, 24, 128)  73728       batch_normalization_18[0][0]     \n",
            "__________________________________________________________________________________________________\n",
            "conv2d_18 (Conv2D)              (None, 24, 24, 128)  409728      p_re_lu_23[0][0]                 \n",
            "__________________________________________________________________________________________________\n",
            "batch_normalization_19 (BatchNo (None, 24, 24, 128)  512         conv2d_18[0][0]                  \n",
            "__________________________________________________________________________________________________\n",
            "p_re_lu_24 (PReLU)              (None, 24, 24, 128)  73728       batch_normalization_19[0][0]     \n",
            "__________________________________________________________________________________________________\n",
            "conv2d_19 (Conv2D)              (None, 24, 24, 128)  409728      p_re_lu_24[0][0]                 \n",
            "__________________________________________________________________________________________________\n",
            "batch_normalization_20 (BatchNo (None, 24, 24, 128)  512         conv2d_19[0][0]                  \n",
            "__________________________________________________________________________________________________\n",
            "p_re_lu_25 (PReLU)              (None, 24, 24, 128)  73728       batch_normalization_20[0][0]     \n",
            "__________________________________________________________________________________________________\n",
            "add_5 (Add)                     (None, 24, 24, 128)  0           p_re_lu_22[0][0]                 \n",
            "                                                                 p_re_lu_25[0][0]                 \n",
            "__________________________________________________________________________________________________\n",
            "p_re_lu_26 (PReLU)              (None, 24, 24, 128)  73728       add_5[0][0]                      \n",
            "__________________________________________________________________________________________________\n",
            "conv2d_transpose_1 (Conv2DTrans (None, 48, 48, 64)   32832       p_re_lu_26[0][0]                 \n",
            "__________________________________________________________________________________________________\n",
            "batch_normalization_21 (BatchNo (None, 48, 48, 64)   256         conv2d_transpose_1[0][0]         \n",
            "__________________________________________________________________________________________________\n",
            "p_re_lu_27 (PReLU)              (None, 48, 48, 64)   147456      batch_normalization_21[0][0]     \n",
            "__________________________________________________________________________________________________\n",
            "concatenate_1 (Concatenate)     (None, 48, 48, 128)  0           p_re_lu_11[0][0]                 \n",
            "                                                                 p_re_lu_27[0][0]                 \n",
            "__________________________________________________________________________________________________\n",
            "conv2d_20 (Conv2D)              (None, 48, 48, 64)   204864      concatenate_1[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "batch_normalization_22 (BatchNo (None, 48, 48, 64)   256         conv2d_20[0][0]                  \n",
            "__________________________________________________________________________________________________\n",
            "p_re_lu_28 (PReLU)              (None, 48, 48, 64)   147456      batch_normalization_22[0][0]     \n",
            "__________________________________________________________________________________________________\n",
            "conv2d_21 (Conv2D)              (None, 48, 48, 64)   102464      p_re_lu_28[0][0]                 \n",
            "__________________________________________________________________________________________________\n",
            "batch_normalization_23 (BatchNo (None, 48, 48, 64)   256         conv2d_21[0][0]                  \n",
            "__________________________________________________________________________________________________\n",
            "p_re_lu_29 (PReLU)              (None, 48, 48, 64)   147456      batch_normalization_23[0][0]     \n",
            "__________________________________________________________________________________________________\n",
            "conv2d_22 (Conv2D)              (None, 48, 48, 64)   102464      p_re_lu_29[0][0]                 \n",
            "__________________________________________________________________________________________________\n",
            "batch_normalization_24 (BatchNo (None, 48, 48, 64)   256         conv2d_22[0][0]                  \n",
            "__________________________________________________________________________________________________\n",
            "p_re_lu_30 (PReLU)              (None, 48, 48, 64)   147456      batch_normalization_24[0][0]     \n",
            "__________________________________________________________________________________________________\n",
            "add_6 (Add)                     (None, 48, 48, 64)   0           p_re_lu_27[0][0]                 \n",
            "                                                                 p_re_lu_30[0][0]                 \n",
            "__________________________________________________________________________________________________\n",
            "p_re_lu_31 (PReLU)              (None, 48, 48, 64)   147456      add_6[0][0]                      \n",
            "__________________________________________________________________________________________________\n",
            "conv2d_transpose_2 (Conv2DTrans (None, 96, 96, 32)   8224        p_re_lu_31[0][0]                 \n",
            "__________________________________________________________________________________________________\n",
            "batch_normalization_25 (BatchNo (None, 96, 96, 32)   128         conv2d_transpose_2[0][0]         \n",
            "__________________________________________________________________________________________________\n",
            "p_re_lu_32 (PReLU)              (None, 96, 96, 32)   294912      batch_normalization_25[0][0]     \n",
            "__________________________________________________________________________________________________\n",
            "concatenate_2 (Concatenate)     (None, 96, 96, 64)   0           p_re_lu_6[0][0]                  \n",
            "                                                                 p_re_lu_32[0][0]                 \n",
            "__________________________________________________________________________________________________\n",
            "conv2d_23 (Conv2D)              (None, 96, 96, 32)   51232       concatenate_2[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "batch_normalization_26 (BatchNo (None, 96, 96, 32)   128         conv2d_23[0][0]                  \n",
            "__________________________________________________________________________________________________\n",
            "p_re_lu_33 (PReLU)              (None, 96, 96, 32)   294912      batch_normalization_26[0][0]     \n",
            "__________________________________________________________________________________________________\n",
            "conv2d_24 (Conv2D)              (None, 96, 96, 32)   25632       p_re_lu_33[0][0]                 \n",
            "__________________________________________________________________________________________________\n",
            "batch_normalization_27 (BatchNo (None, 96, 96, 32)   128         conv2d_24[0][0]                  \n",
            "__________________________________________________________________________________________________\n",
            "p_re_lu_34 (PReLU)              (None, 96, 96, 32)   294912      batch_normalization_27[0][0]     \n",
            "__________________________________________________________________________________________________\n",
            "add_7 (Add)                     (None, 96, 96, 32)   0           p_re_lu_32[0][0]                 \n",
            "                                                                 p_re_lu_34[0][0]                 \n",
            "__________________________________________________________________________________________________\n",
            "p_re_lu_35 (PReLU)              (None, 96, 96, 32)   294912      add_7[0][0]                      \n",
            "__________________________________________________________________________________________________\n",
            "conv2d_transpose_3 (Conv2DTrans (None, 192, 192, 16) 2064        p_re_lu_35[0][0]                 \n",
            "__________________________________________________________________________________________________\n",
            "batch_normalization_28 (BatchNo (None, 192, 192, 16) 64          conv2d_transpose_3[0][0]         \n",
            "__________________________________________________________________________________________________\n",
            "p_re_lu_36 (PReLU)              (None, 192, 192, 16) 589824      batch_normalization_28[0][0]     \n",
            "__________________________________________________________________________________________________\n",
            "concatenate_3 (Concatenate)     (None, 192, 192, 32) 0           p_re_lu_2[0][0]                  \n",
            "                                                                 p_re_lu_36[0][0]                 \n",
            "__________________________________________________________________________________________________\n",
            "conv2d_25 (Conv2D)              (None, 192, 192, 16) 12816       concatenate_3[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "batch_normalization_29 (BatchNo (None, 192, 192, 16) 64          conv2d_25[0][0]                  \n",
            "__________________________________________________________________________________________________\n",
            "p_re_lu_37 (PReLU)              (None, 192, 192, 16) 589824      batch_normalization_29[0][0]     \n",
            "__________________________________________________________________________________________________\n",
            "add_8 (Add)                     (None, 192, 192, 16) 0           p_re_lu_36[0][0]                 \n",
            "                                                                 p_re_lu_37[0][0]                 \n",
            "__________________________________________________________________________________________________\n",
            "p_re_lu_38 (PReLU)              (None, 192, 192, 16) 589824      add_8[0][0]                      \n",
            "__________________________________________________________________________________________________\n",
            "conv2d_26 (Conv2D)              (None, 192, 192, 4)  68          p_re_lu_38[0][0]                 \n",
            "==================================================================================================\n",
            "Total params: 17,304,708\n",
            "Trainable params: 17,299,076\n",
            "Non-trainable params: 5,632\n",
            "__________________________________________________________________________________________________\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.7/dist-packages/tensorflow/python/keras/optimizer_v2/optimizer_v2.py:375: UserWarning: The `lr` argument is deprecated, use `learning_rate` instead.\n",
            "  \"The `lr` argument is deprecated, use `learning_rate` instead.\")\n"
          ],
          "name": "stderr"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Kb78rAHs-4np"
      },
      "source": [
        "checkpointer = callbacks.ModelCheckpoint(filepath = 'vnet-weights.hdf5',save_best_only=True)\n",
        "training_log = callbacks.TensorBoard(log_dir='./Model_logs')"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "GROyVhav_LYT",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "293ad0ea-059e-445c-82ff-91be66dbeb3a"
      },
      "source": [
        "history = model.fit(X_train,Y_train,validation_data=(X_val,Y_val),batch_size=8,epochs=30,callbacks=[training_log,checkpointer],shuffle=True)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.7/dist-packages/keras/utils/generic_utils.py:497: CustomMaskWarning: Custom mask layers require a config and must override get_config. When loading, the custom mask layer must be passed to the custom_objects argument.\n",
            "  category=CustomMaskWarning)\n"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "Epoch 1/30\n",
            "68/68 [==============================] - 83s 291ms/step - loss: 0.3468 - dice_coef: 0.6532 - accuracy: 0.6863 - val_loss: 0.3244 - val_dice_coef: 0.6756 - val_accuracy: 0.9016\n",
            "Epoch 2/30\n",
            "68/68 [==============================] - 13s 193ms/step - loss: 0.0869 - dice_coef: 0.9131 - accuracy: 0.9539 - val_loss: 0.1316 - val_dice_coef: 0.8684 - val_accuracy: 0.9700\n",
            "Epoch 3/30\n",
            "68/68 [==============================] - 13s 195ms/step - loss: 0.0460 - dice_coef: 0.9540 - accuracy: 0.9718 - val_loss: 0.0860 - val_dice_coef: 0.9140 - val_accuracy: 0.9695\n",
            "Epoch 4/30\n",
            "68/68 [==============================] - 13s 197ms/step - loss: 0.0357 - dice_coef: 0.9643 - accuracy: 0.9744 - val_loss: 0.0552 - val_dice_coef: 0.9448 - val_accuracy: 0.9713\n",
            "Epoch 5/30\n",
            "68/68 [==============================] - 13s 198ms/step - loss: 0.0293 - dice_coef: 0.9707 - accuracy: 0.9775 - val_loss: 0.0370 - val_dice_coef: 0.9630 - val_accuracy: 0.9757\n",
            "Epoch 6/30\n",
            "68/68 [==============================] - 13s 196ms/step - loss: 0.0252 - dice_coef: 0.9748 - accuracy: 0.9797 - val_loss: 0.0267 - val_dice_coef: 0.9733 - val_accuracy: 0.9810\n",
            "Epoch 7/30\n",
            "68/68 [==============================] - 13s 195ms/step - loss: 0.0231 - dice_coef: 0.9769 - accuracy: 0.9807 - val_loss: 0.0216 - val_dice_coef: 0.9784 - val_accuracy: 0.9810\n",
            "Epoch 8/30\n",
            "68/68 [==============================] - 13s 196ms/step - loss: 0.0215 - dice_coef: 0.9785 - accuracy: 0.9813 - val_loss: 0.0169 - val_dice_coef: 0.9831 - val_accuracy: 0.9866\n",
            "Epoch 9/30\n",
            "68/68 [==============================] - 13s 197ms/step - loss: 0.0203 - dice_coef: 0.9797 - accuracy: 0.9820 - val_loss: 0.0165 - val_dice_coef: 0.9835 - val_accuracy: 0.9865\n",
            "Epoch 10/30\n",
            "68/68 [==============================] - 13s 197ms/step - loss: 0.0185 - dice_coef: 0.9815 - accuracy: 0.9835 - val_loss: 0.0375 - val_dice_coef: 0.9625 - val_accuracy: 0.9642\n",
            "Epoch 11/30\n",
            "68/68 [==============================] - 13s 196ms/step - loss: 0.0208 - dice_coef: 0.9792 - accuracy: 0.9805 - val_loss: 0.0163 - val_dice_coef: 0.9837 - val_accuracy: 0.9855\n",
            "Epoch 12/30\n",
            "68/68 [==============================] - 13s 196ms/step - loss: 0.0164 - dice_coef: 0.9836 - accuracy: 0.9850 - val_loss: 0.0145 - val_dice_coef: 0.9855 - val_accuracy: 0.9873\n",
            "Epoch 13/30\n",
            "68/68 [==============================] - 13s 195ms/step - loss: 0.0158 - dice_coef: 0.9842 - accuracy: 0.9853 - val_loss: 0.0125 - val_dice_coef: 0.9875 - val_accuracy: 0.9891\n",
            "Epoch 14/30\n",
            "68/68 [==============================] - 13s 195ms/step - loss: 0.0157 - dice_coef: 0.9843 - accuracy: 0.9853 - val_loss: 0.0126 - val_dice_coef: 0.9874 - val_accuracy: 0.9885\n",
            "Epoch 15/30\n",
            "68/68 [==============================] - 13s 195ms/step - loss: 0.0141 - dice_coef: 0.9859 - accuracy: 0.9868 - val_loss: 0.0114 - val_dice_coef: 0.9886 - val_accuracy: 0.9894\n",
            "Epoch 16/30\n",
            "68/68 [==============================] - 13s 196ms/step - loss: 0.0131 - dice_coef: 0.9869 - accuracy: 0.9876 - val_loss: 0.0122 - val_dice_coef: 0.9878 - val_accuracy: 0.9888\n",
            "Epoch 17/30\n",
            "68/68 [==============================] - 13s 195ms/step - loss: 0.0138 - dice_coef: 0.9862 - accuracy: 0.9867 - val_loss: 0.0115 - val_dice_coef: 0.9885 - val_accuracy: 0.9893\n",
            "Epoch 18/30\n",
            "68/68 [==============================] - 13s 195ms/step - loss: 0.0131 - dice_coef: 0.9869 - accuracy: 0.9874 - val_loss: 0.0105 - val_dice_coef: 0.9895 - val_accuracy: 0.9901\n",
            "Epoch 19/30\n",
            "68/68 [==============================] - 13s 196ms/step - loss: 0.0126 - dice_coef: 0.9874 - accuracy: 0.9877 - val_loss: 0.0151 - val_dice_coef: 0.9849 - val_accuracy: 0.9854\n",
            "Epoch 20/30\n",
            "68/68 [==============================] - 13s 195ms/step - loss: 0.0134 - dice_coef: 0.9866 - accuracy: 0.9869 - val_loss: 0.0101 - val_dice_coef: 0.9899 - val_accuracy: 0.9904\n",
            "Epoch 21/30\n",
            "68/68 [==============================] - 13s 196ms/step - loss: 0.0124 - dice_coef: 0.9876 - accuracy: 0.9878 - val_loss: 0.0101 - val_dice_coef: 0.9899 - val_accuracy: 0.9901\n",
            "Epoch 22/30\n",
            "68/68 [==============================] - 13s 196ms/step - loss: 0.0121 - dice_coef: 0.9879 - accuracy: 0.9881 - val_loss: 0.0129 - val_dice_coef: 0.9871 - val_accuracy: 0.9874\n",
            "Epoch 23/30\n",
            "68/68 [==============================] - 13s 196ms/step - loss: 0.0116 - dice_coef: 0.9884 - accuracy: 0.9885 - val_loss: 0.0093 - val_dice_coef: 0.9907 - val_accuracy: 0.9911\n",
            "Epoch 24/30\n",
            "68/68 [==============================] - 13s 196ms/step - loss: 0.0108 - dice_coef: 0.9892 - accuracy: 0.9893 - val_loss: 0.0087 - val_dice_coef: 0.9913 - val_accuracy: 0.9916\n",
            "Epoch 25/30\n",
            "68/68 [==============================] - 13s 197ms/step - loss: 0.0104 - dice_coef: 0.9896 - accuracy: 0.9897 - val_loss: 0.0086 - val_dice_coef: 0.9914 - val_accuracy: 0.9915\n",
            "Epoch 26/30\n",
            "68/68 [==============================] - 13s 197ms/step - loss: 0.0097 - dice_coef: 0.9903 - accuracy: 0.9903 - val_loss: 0.0128 - val_dice_coef: 0.9872 - val_accuracy: 0.9867\n",
            "Epoch 27/30\n",
            "68/68 [==============================] - 13s 197ms/step - loss: 0.0110 - dice_coef: 0.9890 - accuracy: 0.9890 - val_loss: 0.0105 - val_dice_coef: 0.9895 - val_accuracy: 0.9895\n",
            "Epoch 28/30\n",
            "68/68 [==============================] - 13s 196ms/step - loss: 0.0114 - dice_coef: 0.9886 - accuracy: 0.9886 - val_loss: 0.0083 - val_dice_coef: 0.9917 - val_accuracy: 0.9918\n",
            "Epoch 29/30\n",
            "68/68 [==============================] - 13s 196ms/step - loss: 0.0106 - dice_coef: 0.9894 - accuracy: 0.9894 - val_loss: 0.0081 - val_dice_coef: 0.9919 - val_accuracy: 0.9919\n",
            "Epoch 30/30\n",
            "68/68 [==============================] - 13s 196ms/step - loss: 0.0093 - dice_coef: 0.9907 - accuracy: 0.9907 - val_loss: 0.0083 - val_dice_coef: 0.9917 - val_accuracy: 0.9917\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "UDwz_1rC_UwW",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "e53adf8b-0752-45fd-e8f8-397d0ff3b3aa"
      },
      "source": [
        "a = history.history\n",
        "\n",
        "list_traindice = a['dice_coef']\n",
        "list_testdice = a['val_dice_coef']\n",
        "\n",
        "list_trainloss = a['loss']\n",
        "list_testloss = a['val_loss']\n",
        "\n",
        "plt.figure(1)\n",
        "plt.plot(list_testloss, 'b-')\n",
        "plt.plot(list_trainloss,'r-')\n",
        "plt.xlabel('iteration')\n",
        "plt.ylabel('loss')\n",
        "plt.title('loss graph', fontsize = 15)\n",
        "plt.figure(2)\n",
        "plt.plot(list_traindice, 'r-')\n",
        "plt.plot(list_testdice, 'b-')\n",
        "plt.xlabel('iteration')\n",
        "plt.ylabel('accuracy')\n",
        "plt.title('accuracy graph', fontsize = 15)\n",
        "plt.show()\n",
        "\n"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "uW03qQ1x_Yp8"
      },
      "source": [
        "Y_pre = np.argmax(model.predict(X_test),axis=-1)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "NZhdXMGnRvAU",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "cb6a5a2f-443f-4ed1-ff5a-d9e0a8055f00"
      },
      "source": [
        "np.unique(Y_pre)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([0, 2, 3])"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 27
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "oiwctTbhOiG8",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "f885a167-0ede-41a3-d7cd-81dcfca74e94"
      },
      "source": [
        "Y_pre.shape"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(180, 192, 192)"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 28
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "N0j37VFT6VsP"
      },
      "source": [
        "Y_pre=Y_pre.reshape(-1,192,192,1)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "XFvdempo29ff",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "e839288a-038b-45a8-be61-8648e7e1da86"
      },
      "source": [
        "Y_test.shape,X_test.shape"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "((180, 192, 192, 1), (180, 192, 192, 4))"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 30
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "MgS_tGLr_q9h",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        },
        "outputId": "177e3ece-5e4b-4305-baa4-d2cbda3c076d"
      },
      "source": [
        "for i in range(5):\n",
        "  print('X_test '+ str(i))\n",
        "  plt.imshow(X_test[i,:,:,2])\n",
        "  plt.show()\n",
        "\n",
        "  print('Predicted '+ str(i))\n",
        "  plt.imshow(Y_pre[i,:,:,0])\n",
        "  plt.show()\n",
        "\n",
        "  print('Actual '+ str(i))\n",
        "  plt.imshow(Y_test[i,:,:,0])\n",
        "  plt.show()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "X_test 0\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        },
        {
          "output_type": "stream",
          "text": [
            "Predicted 0\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        },
        {
          "output_type": "stream",
          "text": [
            "Actual 0\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        },
        {
          "output_type": "stream",
          "text": [
            "X_test 1\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        },
        {
          "output_type": "stream",
          "text": [
            "Predicted 1\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        },
        {
          "output_type": "stream",
          "text": [
            "Actual 1\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        },
        {
          "output_type": "stream",
          "text": [
            "X_test 2\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        },
        {
          "output_type": "stream",
          "text": [
            "Predicted 2\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        },
        {
          "output_type": "stream",
          "text": [
            "Actual 2\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        },
        {
          "output_type": "stream",
          "text": [
            "X_test 3\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        },
        {
          "output_type": "stream",
          "text": [
            "Predicted 3\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        },
        {
          "output_type": "stream",
          "text": [
            "Actual 3\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAQEAAAD8CAYAAAB3lxGOAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjIsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+WH4yJAAASYElEQVR4nO3de5BcdZnG8e8zk8zkRpjcQQgkxEAV6DrGCMHbgkHElBKwNCa7anRZAyWpvbirRdwqZd2yyloEVsoVK5QUsVQwC4vgEpXLrlqWZkmCVCCBQK6biWFyg2RIwiQz/e4ffRI6k+lkMt093enf86lKTffvnNPnPQ4+c259XkUEZpauhmoXYGbV5RAwS5xDwCxxDgGzxDkEzBLnEDBLXMVCQNI1ktZJWi/plkqtx8xKo0rcJyCpEXgJ+BDQBqwA5kXE2rKvzMxKUqk9gUuB9RGxMSIOAQ8Asyu0LjMrwaAKfe45wNaC923AZcVmblJzDGF4hUoxM4AOXt0VEeN6jlcqBE5K0gJgAcAQhnGZZlarFLMkPBkPbultvFKHA9uAiQXvz83GjoqIxRExPSKmD6a5QmWY2clUKgRWAFMlTZbUBMwFHq3QusysBBU5HIiILkkLgV8BjcC9EbGmEusys9JU7JxARCwDllXq882sPHzHoFniHAJmiXMImCXOIWCWOIeAWeIcAmaJcwiYJc4hYJY4h4BZ4hwCZolzCJglziFgljiHgFniHAJmiXMImCXOIWCWOIeAWeIcAmaJcwiYJc4hYJa4foeApImS/kfSWklrJP1tNn6rpG2Sns3+zSpfuWZWbqU8bbgL+IeIeEbSGcAqSU9k0+6MiG+XXp6ZVVq/QyAitgPbs9cdkl4g34PQzE4jZTknIGkS8E7gf7OhhZJWS7pX0qgiyyyQtFLSysN0lqMMM+uHkkNA0gjgIeDvImIfcDcwBWglv6dwe2/LuRehWW0oKQQkDSYfAD+OiP8EiIj2iOiOiBxwD3Bp6WWaWaWUcnVAwA+AFyLijoLxswtmux54vv/lmVmllXJ14L3AZ4DnJD2bjX0VmCepFQhgM3BjSRWaWUWVcnXgd4B6meQmpGanEd8xaJY4h4BZ4hwCZolzCJglziFgljiHgFniHAJmiXMImCXOIWCWOIeAWeIcAmaJcwiYJc4hYJY4h4BZ4hwCZolzCJglziFgljiHgFniHAJmiSvlQaMASNoMdADdQFdETJc0GvgpMIn8w0bnRMSrpa7LzMqvXHsCV0ZEa0RMz97fAjwVEVOBp7L3ZlaDKnU4MBtYkr1eAlxXofWYWYnKEQIBPC5plaQF2diErGEpwCvAhDKsx8wqoORzAsD7ImKbpPHAE5JeLJwYESEpei6UBcYCgCEMK0MZZtYfJe8JRMS27OcO4GHyvQfbj7Qjy37u6GU5NyQ1qwGlNiQdLumMI6+Bq8n3HnwUmJ/NNh94pJT1mFnllHo4MAF4ON+blEHATyLil5JWAEsl3QBsAeaUuB4zq5CSQiAiNgLv6GV8NzCzlM82s4HhOwbNEucQMEucQ8AscQ4Bs8Q5BMwS5xAwS5xDwCxxDgGzxDkEzBLnEDBLnEPALHEOAbPEOQTMEucQMEucQ8AscQ4Bs8Q5BMwS5xAwS5xDwCxx/X7GoKSLyPcbPOIC4GtAC/AFYGc2/tWIWNbvCs2sovodAhGxDmgFkNQIbCPfd+DzwJ0R8e2yVGhmFVWuw4GZwIaI2FKmzzOzAVKuEJgL3F/wfqGk1ZLulTSqTOswswooOQQkNQHXAv+RDd0NTCF/qLAduL3IcgskrZS08jCdpZZhZv1Ujj2BjwDPREQ7QES0R0R3ROSAe8j3JjyOexGa1YZyhMA8Cg4FjjQizVxPvjehmdWoktqQZU1IPwTcWDD8r5JagQA295hmZjWm1F6E+4ExPcY+U1JFZjagfMegWeIcAmaJcwiYJc4hYJY4h4BZ4hwCZolzCJglziFgljiHgFniHAJmiXMImCXOIWCWOIeAWeIcAmaJcwiYJc4hYJY4h4BZ4hwCZolzCJglrk8hkDUR2SHp+YKx0ZKekPRy9nNUNi5Jd0lanzUgmVap4s2sdH3dE7gPuKbH2C3AUxExFXgqew/5PgRTs38LyDcjMbMa1acQiIjfAnt6DM8GlmSvlwDXFYz/MPKWAy09ehGYWQ0p5ZzAhIjYnr1+BZiQvT4H2FowX1s2ZmY1qCwnBiMiyDcb6TP3IjSrDaWEQPuR3fzs545sfBswsWC+c7OxY7gXoVltKCUEHgXmZ6/nA48UjH82u0owA9hbcNhgZjWmT23IJN0PXAGMldQGfB34FrBU0g3AFmBONvsyYBawHjgAfL7MNZtZGfUpBCJiXpFJM3uZN4CbSynKzAaO7xg0S5xDwCxxDgGzxDkEzBLnEDBLnEPALHEOAbPEOQTMEtenm4XMBlLjuHG8/J0Tf/G0+3ADU+c/M0AV1TeHgNWE1z57OTvflX0RdVAwvmX3CefPhXj5OzOKTm88KC645Q/lLLFuOQRsQA06awJtn5rCWd/5/THjByaI8VN39vlzGhSMn7qr6PTOw4P405ffw1tu+33ReSzP5wRsQMWokXDlq7w+ZwZIR8fHrD3MzpfGlm09zYO7GHrFTjo+NYPX58ygceyYsn12vfGegA0oHeyko200Qz+3Cza/DT3zAtHVRfNjKzhr+Ay6Lzz1z9z92ghyOXHW2L1MHpl/Ct6hXCOrNp/HuPn5vYV93W9lxJazady5l64tW0/0ccnxnoANqK7N/8dFi9ayZ+9w9t56kIYLL0CD+ve3aH9nE/s7mxj8wjCGrR7K2cP38aNJv+ZHk37NHec9wrDVQ3n9jfwDa7pu2M1r33iDLfMm0jhhvPcMCij/zd/qGqnRcZmO+1ay1bldP7+QwY05zrx1KKx4ntc/cSmH5r9KY0PuhMvlQhzuakS/HAXA979yF5c2Dz5uvu7IMf2bC+m++lWGNR0+dt2vnsGUz6wu+NDu0jeoxj0ZD66KiOk9xx0CVjVHQgCg8b6xjFi6HL377bz2zweLLtPefiaj/9BEd7N4dtH3+rSemWuvZf+hphPOM/pvuulev6nvxZ+GioWADwesasb/xXb27B1+7OAJ/ijt2DCGUU83cXC8ePwrt/V5Pe1Pnnv8enrYettQuq9Is0+OTwxa1eQ6OojIXyFov+4N/nTVu1FzN+N5o9f51S32Tg0Wz17M+MYT/5+60B1/fQ93tV3FrgPFlxnefIhoHHpqG1AnHAJWE8a2vA4txafv2DiGQRMO8qV3PMnMoad2/H71sMPcuOEcGpq6GTemo8RK648PB+y00LSngTNGHOSmluOeXn9Ss9bNYvjLTeT2H3/ysND2y5vRu9/e3xJPWycNgSLNSG+T9GLWcPRhSS3Z+CRJByU9m/37fiWLN+uLrY9NonnPyU+Aj3zPDna1jhiAimpLX/YE7uP4ZqRPAG+LiD8DXgIWFUzbEBGt2b+bylOmWWm6hgoGnfjSY6pOGgK9NSONiMcjoit7u5x8lyGziokGyOUaOJA7dErL7c0dRAH7LzvAhLe8VqHqTm/lOCfwV8AvCt5PlvRHSb+R9P5iC7kXoZ2KUdN2sm9TC5f8fGGfl+mOHB/8ly8xuKP698LUspKuDkj6J6AL+HE2tB04LyJ2S3oX8DNJl0TEvp7LRsRiYDHkbxYqpQ5Lw5ipu9m5YyTv/OYX6W6G1f94/M1CN7ZdztNL3nn0fUMEndfsY+wQ/6Eppt8hIOlzwEeBmVnXISKiE/J/1iNilaQNwIXAytJLtdQ1KBg9toODV3aSy4lLvvvF4+Zp7IQYBt2Xvfl3Z0Rz3w8hDny4g86WtL6C3K8QkHQN8BXgzyPiQMH4OGBPRHRLugCYCmwsS6VmwODGHIOHvUEuxK6Ler/k1zA4x9hhvd9wdFJ/HMnEx3ZR/98keNNJQ6BIM9JFQDPwhPLfCV+eXQn4APANSYeBHHBTROzp9YPNgPE/G8L2jzYwfuxxR4wn1KBg/LhTW6Yvhu0Iute+VPbPrWUnDYEizUh/UGTeh4CHSi3K0jFi6XJ47wwo3/NE+m3ni2M5f9OpXX2oB75t2Axo/1MLFy49AMtXn3zmOuMQsKTt3Z//0tDF39xJ16YtVa6mOhwCVnOkOPrtwt7kTjDtVBzuauS8Tz4H5K9zp8ohYDXlrS27uGPiY1z73Pxep+9ZPY7Ji5YPcFX1zSFgNeELk3/HJ87YRCNiRMNw/uvtS/hojyA4+Ph4pvzgOXI18DSseuIQsKra+K3L+furlvHxERs5s2HY0fFRDcc+4KP7wXGc+4uNdHX4eQDl5hCwqmlb9B5u/tgyPj3yBUY1vhkALx3ezyf/7cuM3PzmLTtjVm6la/sr1Siz7jkEbMBpcBOvzZnG8Pft5NMj1zCq4FFhu7r3c9eOD3LWncfetpvyibtKcwjYgGsYOgT+chcNwMrO0bQ0HL3znP9+fRpP/mIa5+M+ggPFIWBVdetLHzv6en9nE4fWnMmkrzkABpJDwAZcRHCoq/G4cf2uhUm3p/PtvVrhELABl+voYNy166pdhmX8tGGzxDkEzBLnEDBLnEPALHEOAbPEOQTMEucQMEtcf3sR3ippW0HPwVkF0xZJWi9pnaQPV6pwMyuP/vYiBLizoOfgMgBJFwNzgUuyZb4n6fhbw8ysZvSrF+EJzAYeiIjOiNgErAcuLaE+M6uwUs4JLMxak98raVQ2dg6wtWCetmzMzGpUf0PgbmAK0Eq+/+Dtp/oBbkhqVhv6FQIR0R4R3RGRA+7hzV3+bcDEglnPzcZ6+4zFETE9IqYPprk/ZZhZGfQrBCSdXfD2euDIlYNHgbmSmiVNJt+L8OnSSjSzSupvL8IrJLUCAWwGbgSIiDWSlgJryT8R6uaISKm3o9lpR1EDj28eqdFxmWZWuwyzuvZkPLgqIqb3HPcdg2aJcwiYJc4hYJY4h4BZ4hwCZolzCJglziFgljiHgFniHAJmiXMImCXOIWCWOIeAWeIcAmaJcwiYJc4hYJY4h4BZ4hwCZolzCJglziFglrj+9iL8aUEfws2Sns3GJ0k6WDDt+5Us3sxKd9KnDZPvRfhd4IdHBiLiU0deS7od2Fsw/4aIaC1XgWZWWScNgYj4raRJvU2TJGAO8MHylmVmA6XUcwLvB9oj4uWCscmS/ijpN5LeX+Lnm1mF9eVw4ETmAfcXvN8OnBcRuyW9C/iZpEsiYl/PBSUtABYADGFYiWWYWX/1e09A0iDg48BPj4xlLcl3Z69XARuAC3tb3r0IzWpDKYcDVwEvRkTbkQFJ4yQ1Zq8vIN+LcGNpJZpZJfXlEuH9wB+AiyS1SbohmzSXYw8FAD4ArM4uGT4I3BQRe8pZsJmVV1+uDswrMv65XsYeAh4qvSwzGyi+Y9AscQ4Bs8Q5BMwS5xAwS5xDwCxxDgGzxDkEzBLnEDBLnEPALHEOAbPEOQTMEucQMEucQ8AscQ4Bs8QpIqpdA5J2AvuBXdWupcLGUt/bWO/bB6f3Np4fEeN6DtZECABIWhkR06tdRyXV+zbW+/ZBfW6jDwfMEucQMEtcLYXA4moXMADqfRvrffugDrexZs4JmFl11NKegJlVQdVDQNI1ktZJWi/plmrXUy5Zt+bnsu7MK7Ox0ZKekPRy9nNUtes8FUU6VPe6Tcq7K/u9rpY0rXqV902R7btV0raCTtuzCqYtyrZvnaQPV6fq0lU1BLJGJf8OfAS4GJgn6eJq1lRmV0ZEa8ElpVuApyJiKvBU9v50ch9wTY+xYtv0EfLNZ6aSbzd39wDVWIr7OH77AO7Mfo+tEbEMIPvvdC5wSbbM94403jndVHtP4FJgfURsjIhDwAPA7CrXVEmzgSXZ6yXAdVWs5ZRFxG+Bns1kim3TbOCHkbccaJF09sBU2j9Ftq+Y2cADWeu9TcB68v89n3aqHQLnAFsL3rdlY/UggMclrcqarwJMiIjt2etXgAnVKa2sim1TPf1uF2aHNPcWHMLVzfZVOwTq2fsiYhr53eKbJX2gcGLkL8vU1aWZetwm8ocxU4BW8l23b69uOeVX7RDYBkwseH9uNnbai4ht2c8dwMPkdxXbj+wSZz93VK/Csim2TXXxu42I9ojojogccA9v7vLXxfZB9UNgBTBV0mRJTeRPtDxa5ZpKJmm4pDOOvAauBp4nv23zs9nmA49Up8KyKrZNjwKfza4SzAD2Fhw2nDZ6nMe4nvzvEfLbN1dSs6TJ5E+APj3Q9ZXDSRuSVlJEdElaCPwKaATujYg11aypTCYAD0uC/P/GP4mIX0paASzNOjtvAeZUscZTlnWovgIYK6kN+DrwLXrfpmXALPInzA4Anx/wgk9Rke27QlIr+cOczcCNABGxRtJSYC3QBdwcEd3VqLtUvmPQLHHVPhwwsypzCJglziFgljiHgFniHAJmiXMImCXOIWCWOIeAWeL+H8UhgzKMmURvAAAAAElFTkSuQmCC\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        },
        {
          "output_type": "stream",
          "text": [
            "X_test 4\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        },
        {
          "output_type": "stream",
          "text": [
            "Predicted 4\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        },
        {
          "output_type": "stream",
          "text": [
            "Actual 4\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "qSovRKb5ErZZ"
      },
      "source": [
        "Y_test_encod = to_categorical(Y_test)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "N_zh-BuKCv4I"
      },
      "source": [
        "train_result = model.evaluate(X_train,Y_train,verbose=0)\n",
        "test_result = model.evaluate(X_test,Y_test_encod,verbose=0)\n",
        "val_result = model.evaluate(X_val,Y_val,verbose=0)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "SwaXWTEFnjIx",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "f240b54b-c23e-4021-dc97-a2a1fe5e7122"
      },
      "source": [
        "cols = ['Loss', 'dice_coef', 'Accuracy']\n",
        "rows = ['Train Result','Validation Result', 'Test Result']\n",
        "\n",
        "cell_text = [train_result, val_result, test_result]\n",
        "\n",
        "rcolors = plt.cm.BuPu(np.full(len(rows), 0.1))\n",
        "ccolors = plt.cm.BuPu(np.full(len(cols), 0.1))\n",
        "\n",
        "plt.figure(linewidth=4,\n",
        "           facecolor= 'white'\n",
        "          )\n",
        "\n",
        "the_table = plt.table(cellText=cell_text,\n",
        "                      rowLabels=rows,\n",
        "                      rowColours=rcolors,\n",
        "                      colColours=ccolors,\n",
        "                      colLabels=cols,\n",
        "                      loc='center')\n",
        "\n",
        "the_table.scale(1.5, 2.5)\n",
        "\n",
        "plt.suptitle('Results of HGG', fontsize=20)\n",
        "\n",
        "ax = plt.gca()\n",
        "ax.get_xaxis().set_visible(False)\n",
        "ax.get_yaxis().set_visible(False)\n",
        "plt.box(on=None)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "pHHZlzMcozLP",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "a7ef5e83-59a8-4cb1-9c36-a6c6d9660ce3"
      },
      "source": [
        "model.save('vnet-model.h5',overwrite=True)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.7/dist-packages/keras/utils/generic_utils.py:497: CustomMaskWarning: Custom mask layers require a config and must override get_config. When loading, the custom mask layer must be passed to the custom_objects argument.\n",
            "  category=CustomMaskWarning)\n"
          ],
          "name": "stderr"
        }
      ]
    }
  ]
}