From 6eee290cb9ee7b8f9bcbad74f630ee5067417aea Mon Sep 17 00:00:00 2001
From: Behrooz <3968947+drbeh@users.noreply.github.com>
Date: Mon, 23 May 2022 14:59:36 +0000
Subject: [PATCH 01/18] Update mil pipeline

Signed-off-by: Behrooz <3968947+drbeh@users.noreply.github.com>
---
 .../panda_mil_train_evaluate_pytorch_gpu.py   | 212 +++++++++---------
 1 file changed, 109 insertions(+), 103 deletions(-)

diff --git a/pathology/multiple_instance_learning/panda_mil_train_evaluate_pytorch_gpu.py b/pathology/multiple_instance_learning/panda_mil_train_evaluate_pytorch_gpu.py
index c37bd5196d..45da2ee931 100644
--- a/pathology/multiple_instance_learning/panda_mil_train_evaluate_pytorch_gpu.py
+++ b/pathology/multiple_instance_learning/panda_mil_train_evaluate_pytorch_gpu.py
@@ -20,74 +20,23 @@
 import torch.multiprocessing as mp
 
 from monai.data import Dataset, load_decathlon_datalist
-from monai.data.image_reader import WSIReader
+from monai.data.wsi_reader import WSIReader
 from monai.metrics import Cumulative, CumulativeAverage
-from monai.transforms import Transform, Compose, LoadImageD, RandFlipd, RandRotate90d, ScaleIntensityRangeD, ToTensord
-from monai.apps.pathology.transforms import TileOnGridd
+from monai.transforms import (
+    Transform,
+    Compose,
+    LoadImaged,
+    RandFlipd,
+    GridPatchd,
+    RandRotate90d,
+    ScaleIntensityRanged,
+    ToTensord,
+)
+from monai.transforms.spatial.array import GridPatch
+from monai.transforms.spatial.dictionary import GridPatchd
 from monai.networks.nets import milmodel
 
 
-def parse_args():
-
-    parser = argparse.ArgumentParser(description="Multiple Instance Learning (MIL) example of classification from WSI.")
-    parser.add_argument(
-        "--data_root", default="/PandaChallenge2020/train_images/", help="path to root folder of images"
-    )
-    parser.add_argument("--dataset_json", default=None, type=str, help="path to dataset json file")
-
-    parser.add_argument("--num_classes", default=5, type=int, help="number of output classes")
-    parser.add_argument("--mil_mode", default="att_trans", help="MIL algorithm")
-    parser.add_argument(
-        "--tile_count", default=44, type=int, help="number of patches (instances) to extract from WSI image"
-    )
-    parser.add_argument("--tile_size", default=256, type=int, help="size of square patch (instance) in pixels")
-
-    parser.add_argument("--checkpoint", default=None, help="load existing checkpoint")
-    parser.add_argument(
-        "--validate",
-        action="store_true",
-        help="run only inference on the validation set, must specify the checkpoint argument",
-    )
-
-    parser.add_argument("--logdir", default=None, help="path to log directory to store Tensorboard logs")
-
-    parser.add_argument("--epochs", default=50, type=int, help="number of training epochs")
-    parser.add_argument("--batch_size", default=4, type=int, help="batch size, the number of WSI images per gpu")
-    parser.add_argument("--optim_lr", default=3e-5, type=float, help="initial learning rate")
-
-    parser.add_argument("--weight_decay", default=0, type=float, help="optimizer weight decay")
-    parser.add_argument("--amp", action="store_true", help="use AMP, recommended")
-    parser.add_argument(
-        "--val_every",
-        default=1,
-        type=int,
-        help="run validation after this number of epochs, default 1 to run every epoch",
-    )
-    parser.add_argument("--workers", default=2, type=int, help="number of workers for data loading")
-
-    ###for multigpu
-    parser.add_argument("--distributed", action="store_true", help="use multigpu training, recommended")
-    parser.add_argument("--world_size", default=1, type=int, help="number of nodes for distributed training")
-    parser.add_argument("--rank", default=0, type=int, help="node rank for distributed training")
-    parser.add_argument(
-        "--dist-url", default="tcp://127.0.0.1:23456", type=str, help="url used to set up distributed training"
-    )
-    parser.add_argument("--dist-backend", default="nccl", type=str, help="distributed backend")
-
-    parser.add_argument(
-        "--quick", action="store_true", help="use a small subset of data for debugging"
-    )  # for debugging
-
-    args = parser.parse_args()
-
-    print("Argument values:")
-    for k, v in vars(args).items():
-        print(k, "=>", v)
-    print("-----------------")
-
-    return args
-
-
 def train_epoch(model, loader, optimizer, scaler, epoch, args):
     """One train epoch over the dataset"""
 
@@ -278,35 +227,12 @@ def __call__(self, data):
         return d
 
 
-def main():
-
-    args = parse_args()
-
-    if args.dataset_json is None:
-        # download default json datalist
-        resource = "https://drive.google.com/uc?id=1L6PtKBlHHyUgTE4rVhRuOLTQKgD4tBRK"
-        dst = "./datalist_panda_0.json"
-        if not os.path.exists(dst):
-            gdown.download(resource, dst, quiet=False)
-        args.dataset_json = dst
-
-    if args.distributed:
-        ngpus_per_node = torch.cuda.device_count()
-        args.optim_lr = ngpus_per_node * args.optim_lr / 2  # heuristic to scale up learning rate in multigpu setup
-        args.world_size = ngpus_per_node * args.world_size
-
-        print("Multigpu", ngpus_per_node, "rescaled lr", args.optim_lr)
-        mp.spawn(main_worker, nprocs=ngpus_per_node, args=(args,))
-    else:
-        main_worker(0, args)
-
-
 def list_data_collate(batch: collections.abc.Sequence):
-    '''
-        Combine instances from a list of dicts into a single dict, by stacking them along first dim
-        [{'image' : 3xHxW}, {'image' : 3xHxW}, {'image' : 3xHxW}...] - > {'image' : Nx3xHxW}
-        followed by the default collate which will form a batch BxNx3xHxW
-    '''
+    """
+    Combine instances from a list of dicts into a single dict, by stacking them along first dim
+    [{'image' : 3xHxW}, {'image' : 3xHxW}, {'image' : 3xHxW}...] - > {'image' : Nx3xHxW}
+    followed by the default collate which will form a batch BxNx3xHxW
+    """
 
     for i, item in enumerate(batch):
         data = item[0]
@@ -352,27 +278,27 @@ def main_worker(gpu, args):
 
     train_transform = Compose(
         [
-            LoadImageD(keys=["image"], reader=WSIReader, backend="TiffFile", dtype=np.uint8, level=1, image_only=True),
+            LoadImaged(keys=["image"], reader=WSIReader, backend="cupy", dtype=np.uint8, level=1, image_only=True),
             LabelEncodeIntegerGraded(keys=["label"], num_classes=args.num_classes),
-            TileOnGridd(
+            GridPatchd(
                 keys=["image"],
-                tile_count=args.tile_count,
-                tile_size=args.tile_size,
-                random_offset=True,
-                background_val=255,
-                return_list_of_dicts=True,
+                patch_size=args.tile_size,
+                start_pos="random",
+                max_num_patches=args.tile_count,
+                overlap=0.0,
+                pad_opts={"constant_values": 255},
             ),
             RandFlipd(keys=["image"], spatial_axis=0, prob=0.5),
             RandFlipd(keys=["image"], spatial_axis=1, prob=0.5),
             RandRotate90d(keys=["image"], prob=0.5),
-            ScaleIntensityRangeD(keys=["image"], a_min=np.float32(255), a_max=np.float32(0)),
+            ScaleIntensityRanged(keys=["image"], a_min=np.float32(255), a_max=np.float32(0)),
             ToTensord(keys=["image", "label"]),
         ]
     )
 
     valid_transform = Compose(
         [
-            LoadImageD(keys=["image"], reader=WSIReader, backend="TiffFile", dtype=np.uint8, level=1, image_only=True),
+            LoadImaged(keys=["image"], reader=WSIReader, backend="TiffFile", dtype=np.uint8, level=1, image_only=True),
             LabelEncodeIntegerGraded(keys=["label"], num_classes=args.num_classes),
             TileOnGridd(
                 keys=["image"],
@@ -382,7 +308,7 @@ def main_worker(gpu, args):
                 background_val=255,
                 return_list_of_dicts=True,
             ),
-            ScaleIntensityRangeD(keys=["image"], a_min=np.float32(255), a_max=np.float32(0)),
+            ScaleIntensityRanged(keys=["image"], a_min=np.float32(255), a_max=np.float32(0)),
             ToTensord(keys=["image", "label"]),
         ]
     )
@@ -540,5 +466,85 @@ def main_worker(gpu, args):
     print("ALL DONE")
 
 
+def parse_args():
+
+    parser = argparse.ArgumentParser(description="Multiple Instance Learning (MIL) example of classification from WSI.")
+    parser.add_argument(
+        "--data_root", default="/PandaChallenge2020/train_images/", help="path to root folder of images"
+    )
+    parser.add_argument("--dataset_json", default=None, type=str, help="path to dataset json file")
+
+    parser.add_argument("--num_classes", default=5, type=int, help="number of output classes")
+    parser.add_argument("--mil_mode", default="att_trans", help="MIL algorithm")
+    parser.add_argument(
+        "--tile_count", default=44, type=int, help="number of patches (instances) to extract from WSI image"
+    )
+    parser.add_argument("--tile_size", default=256, type=int, help="size of square patch (instance) in pixels")
+
+    parser.add_argument("--checkpoint", default=None, help="load existing checkpoint")
+    parser.add_argument(
+        "--validate",
+        action="store_true",
+        help="run only inference on the validation set, must specify the checkpoint argument",
+    )
+
+    parser.add_argument("--logdir", default=None, help="path to log directory to store Tensorboard logs")
+
+    parser.add_argument("--epochs", default=50, type=int, help="number of training epochs")
+    parser.add_argument("--batch_size", default=4, type=int, help="batch size, the number of WSI images per gpu")
+    parser.add_argument("--optim_lr", default=3e-5, type=float, help="initial learning rate")
+
+    parser.add_argument("--weight_decay", default=0, type=float, help="optimizer weight decay")
+    parser.add_argument("--amp", action="store_true", help="use AMP, recommended")
+    parser.add_argument(
+        "--val_every",
+        default=1,
+        type=int,
+        help="run validation after this number of epochs, default 1 to run every epoch",
+    )
+    parser.add_argument("--workers", default=2, type=int, help="number of workers for data loading")
+
+    ###for multigpu
+    parser.add_argument("--distributed", action="store_true", help="use multigpu training, recommended")
+    parser.add_argument("--world_size", default=1, type=int, help="number of nodes for distributed training")
+    parser.add_argument("--rank", default=0, type=int, help="node rank for distributed training")
+    parser.add_argument(
+        "--dist-url", default="tcp://127.0.0.1:23456", type=str, help="url used to set up distributed training"
+    )
+    parser.add_argument("--dist-backend", default="nccl", type=str, help="distributed backend")
+
+    parser.add_argument(
+        "--quick", action="store_true", help="use a small subset of data for debugging"
+    )  # for debugging
+
+    args = parser.parse_args()
+
+    print("Argument values:")
+    for k, v in vars(args).items():
+        print(k, "=>", v)
+    print("-----------------")
+
+    return args
+
+
 if __name__ == "__main__":
-    main()
+
+    args = parse_args()
+
+    if args.dataset_json is None:
+        # download default json datalist
+        resource = "https://drive.google.com/uc?id=1L6PtKBlHHyUgTE4rVhRuOLTQKgD4tBRK"
+        dst = "./datalist_panda_0.json"
+        if not os.path.exists(dst):
+            gdown.download(resource, dst, quiet=False)
+        args.dataset_json = dst
+
+    if args.distributed:
+        ngpus_per_node = torch.cuda.device_count()
+        args.optim_lr = ngpus_per_node * args.optim_lr / 2  # heuristic to scale up learning rate in multigpu setup
+        args.world_size = ngpus_per_node * args.world_size
+
+        print("Multigpu", ngpus_per_node, "rescaled lr", args.optim_lr)
+        mp.spawn(main_worker, nprocs=ngpus_per_node, args=(args,))
+    else:
+        main_worker(0, args)

From 53252952c4a9a15abd14b5d534844aa9a6e5f84a Mon Sep 17 00:00:00 2001
From: Behrooz <3968947+drbeh@users.noreply.github.com>
Date: Mon, 23 May 2022 15:01:49 +0000
Subject: [PATCH 02/18] Fix a typo

Signed-off-by: Behrooz <3968947+drbeh@users.noreply.github.com>
---
 .../panda_mil_train_evaluate_pytorch_gpu.py                     | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/pathology/multiple_instance_learning/panda_mil_train_evaluate_pytorch_gpu.py b/pathology/multiple_instance_learning/panda_mil_train_evaluate_pytorch_gpu.py
index 45da2ee931..1ce03f6493 100644
--- a/pathology/multiple_instance_learning/panda_mil_train_evaluate_pytorch_gpu.py
+++ b/pathology/multiple_instance_learning/panda_mil_train_evaluate_pytorch_gpu.py
@@ -278,7 +278,7 @@ def main_worker(gpu, args):
 
     train_transform = Compose(
         [
-            LoadImaged(keys=["image"], reader=WSIReader, backend="cupy", dtype=np.uint8, level=1, image_only=True),
+            LoadImaged(keys=["image"], reader=WSIReader, backend="cucim", dtype=np.uint8, level=1, image_only=True),
             LabelEncodeIntegerGraded(keys=["label"], num_classes=args.num_classes),
             GridPatchd(
                 keys=["image"],

From afb66f731d25057a42ebe0f7e711604883c0689f Mon Sep 17 00:00:00 2001
From: Behrooz <3968947+drbeh@users.noreply.github.com>
Date: Mon, 23 May 2022 15:06:44 +0000
Subject: [PATCH 03/18] Tiffile to cucim

Signed-off-by: Behrooz <3968947+drbeh@users.noreply.github.com>
---
 .../panda_mil_train_evaluate_pytorch_gpu.py                     | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/pathology/multiple_instance_learning/panda_mil_train_evaluate_pytorch_gpu.py b/pathology/multiple_instance_learning/panda_mil_train_evaluate_pytorch_gpu.py
index 1ce03f6493..e01d7e0833 100644
--- a/pathology/multiple_instance_learning/panda_mil_train_evaluate_pytorch_gpu.py
+++ b/pathology/multiple_instance_learning/panda_mil_train_evaluate_pytorch_gpu.py
@@ -298,7 +298,7 @@ def main_worker(gpu, args):
 
     valid_transform = Compose(
         [
-            LoadImaged(keys=["image"], reader=WSIReader, backend="TiffFile", dtype=np.uint8, level=1, image_only=True),
+            LoadImaged(keys=["image"], reader=WSIReader, backend="cucim", dtype=np.uint8, level=1, image_only=True),
             LabelEncodeIntegerGraded(keys=["label"], num_classes=args.num_classes),
             TileOnGridd(
                 keys=["image"],

From de9f7ebab7330a90a1ff79192b8f6cdf197aa8f9 Mon Sep 17 00:00:00 2001
From: Behrooz <3968947+drbeh@users.noreply.github.com>
Date: Mon, 23 May 2022 15:15:05 +0000
Subject: [PATCH 04/18] Update to RandGridPatchd

Signed-off-by: Behrooz <3968947+drbeh@users.noreply.github.com>
---
 .../panda_mil_train_evaluate_pytorch_gpu.py        | 14 +++++---------
 1 file changed, 5 insertions(+), 9 deletions(-)

diff --git a/pathology/multiple_instance_learning/panda_mil_train_evaluate_pytorch_gpu.py b/pathology/multiple_instance_learning/panda_mil_train_evaluate_pytorch_gpu.py
index e01d7e0833..33281cc3b5 100644
--- a/pathology/multiple_instance_learning/panda_mil_train_evaluate_pytorch_gpu.py
+++ b/pathology/multiple_instance_learning/panda_mil_train_evaluate_pytorch_gpu.py
@@ -280,12 +280,10 @@ def main_worker(gpu, args):
         [
             LoadImaged(keys=["image"], reader=WSIReader, backend="cucim", dtype=np.uint8, level=1, image_only=True),
             LabelEncodeIntegerGraded(keys=["label"], num_classes=args.num_classes),
-            GridPatchd(
+            RandGridPatchd(
                 keys=["image"],
                 patch_size=args.tile_size,
-                start_pos="random",
                 max_num_patches=args.tile_count,
-                overlap=0.0,
                 pad_opts={"constant_values": 255},
             ),
             RandFlipd(keys=["image"], spatial_axis=0, prob=0.5),
@@ -300,13 +298,11 @@ def main_worker(gpu, args):
         [
             LoadImaged(keys=["image"], reader=WSIReader, backend="cucim", dtype=np.uint8, level=1, image_only=True),
             LabelEncodeIntegerGraded(keys=["label"], num_classes=args.num_classes),
-            TileOnGridd(
+            GridPatchd(
                 keys=["image"],
-                tile_count=None,
-                tile_size=args.tile_size,
-                random_offset=False,
-                background_val=255,
-                return_list_of_dicts=True,
+                patch_size=args.tile_size,
+                max_num_patches=args.tile_count,
+                pad_opts={"constant_values": 255},
             ),
             ScaleIntensityRanged(keys=["image"], a_min=np.float32(255), a_max=np.float32(0)),
             ToTensord(keys=["image", "label"]),

From a54a17d06d1f154df56c62d52c1af60bdf41d5cd Mon Sep 17 00:00:00 2001
From: Behrooz <3968947+drbeh@users.noreply.github.com>
Date: Mon, 23 May 2022 15:15:59 +0000
Subject: [PATCH 05/18] Fix import

Signed-off-by: Behrooz <3968947+drbeh@users.noreply.github.com>
---
 .../panda_mil_train_evaluate_pytorch_gpu.py                      | 1 +
 1 file changed, 1 insertion(+)

diff --git a/pathology/multiple_instance_learning/panda_mil_train_evaluate_pytorch_gpu.py b/pathology/multiple_instance_learning/panda_mil_train_evaluate_pytorch_gpu.py
index 33281cc3b5..293c1d32ef 100644
--- a/pathology/multiple_instance_learning/panda_mil_train_evaluate_pytorch_gpu.py
+++ b/pathology/multiple_instance_learning/panda_mil_train_evaluate_pytorch_gpu.py
@@ -27,6 +27,7 @@
     Compose,
     LoadImaged,
     RandFlipd,
+    RandGridPatchd,
     GridPatchd,
     RandRotate90d,
     ScaleIntensityRanged,

From 554ed71e3e8e8a4cfbfbcb4077f29bc7f111634c Mon Sep 17 00:00:00 2001
From: Behrooz <3968947+drbeh@users.noreply.github.com>
Date: Mon, 23 May 2022 15:17:22 +0000
Subject: [PATCH 06/18] Add sort_key

Signed-off-by: Behrooz <3968947+drbeh@users.noreply.github.com>
---
 .../panda_mil_train_evaluate_pytorch_gpu.py                     | 2 ++
 1 file changed, 2 insertions(+)

diff --git a/pathology/multiple_instance_learning/panda_mil_train_evaluate_pytorch_gpu.py b/pathology/multiple_instance_learning/panda_mil_train_evaluate_pytorch_gpu.py
index 293c1d32ef..ebd0ee0ee4 100644
--- a/pathology/multiple_instance_learning/panda_mil_train_evaluate_pytorch_gpu.py
+++ b/pathology/multiple_instance_learning/panda_mil_train_evaluate_pytorch_gpu.py
@@ -285,6 +285,7 @@ def main_worker(gpu, args):
                 keys=["image"],
                 patch_size=args.tile_size,
                 max_num_patches=args.tile_count,
+                sort_key="max",
                 pad_opts={"constant_values": 255},
             ),
             RandFlipd(keys=["image"], spatial_axis=0, prob=0.5),
@@ -303,6 +304,7 @@ def main_worker(gpu, args):
                 keys=["image"],
                 patch_size=args.tile_size,
                 max_num_patches=args.tile_count,
+                sort_key="max",
                 pad_opts={"constant_values": 255},
             ),
             ScaleIntensityRanged(keys=["image"], a_min=np.float32(255), a_max=np.float32(0)),

From 0724dfe3b2aa5f283625dd40950c34d8b2f64bb8 Mon Sep 17 00:00:00 2001
From: Behrooz <3968947+drbeh@users.noreply.github.com>
Date: Mon, 23 May 2022 18:23:42 +0000
Subject: [PATCH 07/18] sort key max to min

Signed-off-by: Behrooz <3968947+drbeh@users.noreply.github.com>
---
 .../panda_mil_train_evaluate_pytorch_gpu.py                  | 5 ++---
 1 file changed, 2 insertions(+), 3 deletions(-)

diff --git a/pathology/multiple_instance_learning/panda_mil_train_evaluate_pytorch_gpu.py b/pathology/multiple_instance_learning/panda_mil_train_evaluate_pytorch_gpu.py
index ebd0ee0ee4..a790695651 100644
--- a/pathology/multiple_instance_learning/panda_mil_train_evaluate_pytorch_gpu.py
+++ b/pathology/multiple_instance_learning/panda_mil_train_evaluate_pytorch_gpu.py
@@ -285,7 +285,7 @@ def main_worker(gpu, args):
                 keys=["image"],
                 patch_size=args.tile_size,
                 max_num_patches=args.tile_count,
-                sort_key="max",
+                sort_key="min",
                 pad_opts={"constant_values": 255},
             ),
             RandFlipd(keys=["image"], spatial_axis=0, prob=0.5),
@@ -303,8 +303,7 @@ def main_worker(gpu, args):
             GridPatchd(
                 keys=["image"],
                 patch_size=args.tile_size,
-                max_num_patches=args.tile_count,
-                sort_key="max",
+                sort_key="min",
                 pad_opts={"constant_values": 255},
             ),
             ScaleIntensityRanged(keys=["image"], a_min=np.float32(255), a_max=np.float32(0)),

From 7f34968f8df6903122b9add6f79037a1f18af54f Mon Sep 17 00:00:00 2001
From: Behrooz <3968947+drbeh@users.noreply.github.com>
Date: Mon, 23 May 2022 18:56:03 +0000
Subject: [PATCH 08/18] Fix patch_size

Signed-off-by: Behrooz <3968947+drbeh@users.noreply.github.com>
---
 .../panda_mil_train_evaluate_pytorch_gpu.py                   | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/pathology/multiple_instance_learning/panda_mil_train_evaluate_pytorch_gpu.py b/pathology/multiple_instance_learning/panda_mil_train_evaluate_pytorch_gpu.py
index a790695651..6be73ee933 100644
--- a/pathology/multiple_instance_learning/panda_mil_train_evaluate_pytorch_gpu.py
+++ b/pathology/multiple_instance_learning/panda_mil_train_evaluate_pytorch_gpu.py
@@ -283,7 +283,7 @@ def main_worker(gpu, args):
             LabelEncodeIntegerGraded(keys=["label"], num_classes=args.num_classes),
             RandGridPatchd(
                 keys=["image"],
-                patch_size=args.tile_size,
+                patch_size=(args.tile_size, args.tile_size),
                 max_num_patches=args.tile_count,
                 sort_key="min",
                 pad_opts={"constant_values": 255},
@@ -302,7 +302,7 @@ def main_worker(gpu, args):
             LabelEncodeIntegerGraded(keys=["label"], num_classes=args.num_classes),
             GridPatchd(
                 keys=["image"],
-                patch_size=args.tile_size,
+                patch_size=(args.tile_size, args.tile_size),
                 sort_key="min",
                 pad_opts={"constant_values": 255},
             ),

From a99fe581557b7efe0e68af12b0dd165d5ee887c2 Mon Sep 17 00:00:00 2001
From: Behrooz <3968947+drbeh@users.noreply.github.com>
Date: Mon, 23 May 2022 19:37:42 +0000
Subject: [PATCH 09/18] Stack patch locations

Signed-off-by: Behrooz <3968947+drbeh@users.noreply.github.com>
---
 .../panda_mil_train_evaluate_pytorch_gpu.py                      | 1 +
 1 file changed, 1 insertion(+)

diff --git a/pathology/multiple_instance_learning/panda_mil_train_evaluate_pytorch_gpu.py b/pathology/multiple_instance_learning/panda_mil_train_evaluate_pytorch_gpu.py
index 6be73ee933..953e9a293f 100644
--- a/pathology/multiple_instance_learning/panda_mil_train_evaluate_pytorch_gpu.py
+++ b/pathology/multiple_instance_learning/panda_mil_train_evaluate_pytorch_gpu.py
@@ -238,6 +238,7 @@ def list_data_collate(batch: collections.abc.Sequence):
     for i, item in enumerate(batch):
         data = item[0]
         data["image"] = torch.stack([ix["image"] for ix in item], dim=0)
+        data["patch"]['location'] = [ix["patch"]['location'] for ix in item]
         batch[i] = data
     return default_collate(batch)
 

From 8116255b0adb187897dde98697abaa0c63296cf1 Mon Sep 17 00:00:00 2001
From: Behrooz <3968947+drbeh@users.noreply.github.com>
Date: Fri, 27 May 2022 02:50:37 +0000
Subject: [PATCH 10/18] Remove location

Signed-off-by: Behrooz <3968947+drbeh@users.noreply.github.com>
---
 .../panda_mil_train_evaluate_pytorch_gpu.py                | 7 ++-----
 1 file changed, 2 insertions(+), 5 deletions(-)

diff --git a/pathology/multiple_instance_learning/panda_mil_train_evaluate_pytorch_gpu.py b/pathology/multiple_instance_learning/panda_mil_train_evaluate_pytorch_gpu.py
index 953e9a293f..e65c865a24 100644
--- a/pathology/multiple_instance_learning/panda_mil_train_evaluate_pytorch_gpu.py
+++ b/pathology/multiple_instance_learning/panda_mil_train_evaluate_pytorch_gpu.py
@@ -33,8 +33,6 @@
     ScaleIntensityRanged,
     ToTensord,
 )
-from monai.transforms.spatial.array import GridPatch
-from monai.transforms.spatial.dictionary import GridPatchd
 from monai.networks.nets import milmodel
 
 
@@ -209,7 +207,7 @@ class LabelEncodeIntegerGraded(Transform):
 
     """
 
-    def __init__(self, num_classes, keys=["label"]):
+    def __init__(self, num_classes, keys=("label",)):
         super().__init__()
         self.keys = keys
         self.num_classes = num_classes
@@ -238,7 +236,6 @@ def list_data_collate(batch: collections.abc.Sequence):
     for i, item in enumerate(batch):
         data = item[0]
         data["image"] = torch.stack([ix["image"] for ix in item], dim=0)
-        data["patch"]['location'] = [ix["patch"]['location'] for ix in item]
         batch[i] = data
     return default_collate(batch)
 
@@ -285,7 +282,7 @@ def main_worker(gpu, args):
             RandGridPatchd(
                 keys=["image"],
                 patch_size=(args.tile_size, args.tile_size),
-                max_num_patches=args.tile_count,
+                num_patches=args.tile_count,
                 sort_key="min",
                 pad_opts={"constant_values": 255},
             ),

From a07e1217c4fa31e406778d93801f571cf6b563b6 Mon Sep 17 00:00:00 2001
From: Behrooz <3968947+drbeh@users.noreply.github.com>
Date: Fri, 27 May 2022 12:58:28 +0000
Subject: [PATCH 11/18] Change to fix_num_patches

Signed-off-by: Behrooz <3968947+drbeh@users.noreply.github.com>
---
 .../panda_mil_train_evaluate_pytorch_gpu.py                     | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/pathology/multiple_instance_learning/panda_mil_train_evaluate_pytorch_gpu.py b/pathology/multiple_instance_learning/panda_mil_train_evaluate_pytorch_gpu.py
index e65c865a24..4cf14c882d 100644
--- a/pathology/multiple_instance_learning/panda_mil_train_evaluate_pytorch_gpu.py
+++ b/pathology/multiple_instance_learning/panda_mil_train_evaluate_pytorch_gpu.py
@@ -282,7 +282,7 @@ def main_worker(gpu, args):
             RandGridPatchd(
                 keys=["image"],
                 patch_size=(args.tile_size, args.tile_size),
-                num_patches=args.tile_count,
+                fix_num_patches=args.tile_count,
                 sort_key="min",
                 pad_opts={"constant_values": 255},
             ),

From 975edc4fe68393a7cd4df811751eb716dd6f49f8 Mon Sep 17 00:00:00 2001
From: Behrooz <3968947+drbeh@users.noreply.github.com>
Date: Tue, 31 May 2022 12:34:54 +0000
Subject: [PATCH 12/18] Update grid patch

Signed-off-by: Behrooz <3968947+drbeh@users.noreply.github.com>
---
 .../panda_mil_train_evaluate_pytorch_gpu.py            | 10 +++++-----
 1 file changed, 5 insertions(+), 5 deletions(-)

diff --git a/pathology/multiple_instance_learning/panda_mil_train_evaluate_pytorch_gpu.py b/pathology/multiple_instance_learning/panda_mil_train_evaluate_pytorch_gpu.py
index 4cf14c882d..912177fd6a 100644
--- a/pathology/multiple_instance_learning/panda_mil_train_evaluate_pytorch_gpu.py
+++ b/pathology/multiple_instance_learning/panda_mil_train_evaluate_pytorch_gpu.py
@@ -282,9 +282,9 @@ def main_worker(gpu, args):
             RandGridPatchd(
                 keys=["image"],
                 patch_size=(args.tile_size, args.tile_size),
-                fix_num_patches=args.tile_count,
-                sort_key="min",
-                pad_opts={"constant_values": 255},
+                num_patches=args.tile_count,
+                sort_fn="min",
+                constant_values=255,
             ),
             RandFlipd(keys=["image"], spatial_axis=0, prob=0.5),
             RandFlipd(keys=["image"], spatial_axis=1, prob=0.5),
@@ -301,8 +301,8 @@ def main_worker(gpu, args):
             GridPatchd(
                 keys=["image"],
                 patch_size=(args.tile_size, args.tile_size),
-                sort_key="min",
-                pad_opts={"constant_values": 255},
+                sort_fn="min",
+                constant_values=255,
             ),
             ScaleIntensityRanged(keys=["image"], a_min=np.float32(255), a_max=np.float32(0)),
             ToTensord(keys=["image", "label"]),

From e21b81d3d242c63c948cb1da655fb74e3b898c91 Mon Sep 17 00:00:00 2001
From: Behrooz <3968947+drbeh@users.noreply.github.com>
Date: Thu, 2 Jun 2022 00:02:12 +0000
Subject: [PATCH 13/18] Add threshold_filter

Signed-off-by: Behrooz <3968947+drbeh@users.noreply.github.com>
---
 .../panda_mil_train_evaluate_pytorch_gpu.py                 | 6 ++++++
 1 file changed, 6 insertions(+)

diff --git a/pathology/multiple_instance_learning/panda_mil_train_evaluate_pytorch_gpu.py b/pathology/multiple_instance_learning/panda_mil_train_evaluate_pytorch_gpu.py
index 912177fd6a..8c664fe764 100644
--- a/pathology/multiple_instance_learning/panda_mil_train_evaluate_pytorch_gpu.py
+++ b/pathology/multiple_instance_learning/panda_mil_train_evaluate_pytorch_gpu.py
@@ -275,6 +275,10 @@ def main_worker(gpu, args):
         training_list = training_list[:16]
         validation_list = validation_list[:16]
 
+    def threshold_filter(patch):
+        thresh = 0.999 * 3 * 255 * args.tile_size * args.tile_size
+        return patch.sum() < thresh
+
     train_transform = Compose(
         [
             LoadImaged(keys=["image"], reader=WSIReader, backend="cucim", dtype=np.uint8, level=1, image_only=True),
@@ -284,6 +288,7 @@ def main_worker(gpu, args):
                 patch_size=(args.tile_size, args.tile_size),
                 num_patches=args.tile_count,
                 sort_fn="min",
+                filter_fn=threshold_filter,
                 constant_values=255,
             ),
             RandFlipd(keys=["image"], spatial_axis=0, prob=0.5),
@@ -302,6 +307,7 @@ def main_worker(gpu, args):
                 keys=["image"],
                 patch_size=(args.tile_size, args.tile_size),
                 sort_fn="min",
+                filter_fn=threshold_filter,
                 constant_values=255,
             ),
             ScaleIntensityRanged(keys=["image"], a_min=np.float32(255), a_max=np.float32(0)),

From 10ef5fa680bdf0b308dceb5b65d7c07cd96e8362 Mon Sep 17 00:00:00 2001
From: Behrooz <3968947+drbeh@users.noreply.github.com>
Date: Thu, 2 Jun 2022 00:53:13 +0000
Subject: [PATCH 14/18] Update threshold and add pad_mode=None

Signed-off-by: Behrooz <3968947+drbeh@users.noreply.github.com>
---
 .../panda_mil_train_evaluate_pytorch_gpu.py        | 14 +++++---------
 1 file changed, 5 insertions(+), 9 deletions(-)

diff --git a/pathology/multiple_instance_learning/panda_mil_train_evaluate_pytorch_gpu.py b/pathology/multiple_instance_learning/panda_mil_train_evaluate_pytorch_gpu.py
index 8c664fe764..e0ab543e5e 100644
--- a/pathology/multiple_instance_learning/panda_mil_train_evaluate_pytorch_gpu.py
+++ b/pathology/multiple_instance_learning/panda_mil_train_evaluate_pytorch_gpu.py
@@ -35,7 +35,7 @@
 )
 from monai.networks.nets import milmodel
 
-
+    
 def train_epoch(model, loader, optimizer, scaler, epoch, args):
     """One train epoch over the dataset"""
 
@@ -238,8 +238,7 @@ def list_data_collate(batch: collections.abc.Sequence):
         data["image"] = torch.stack([ix["image"] for ix in item], dim=0)
         batch[i] = data
     return default_collate(batch)
-
-
+    
 def main_worker(gpu, args):
 
     args.gpu = gpu
@@ -275,10 +274,6 @@ def main_worker(gpu, args):
         training_list = training_list[:16]
         validation_list = validation_list[:16]
 
-    def threshold_filter(patch):
-        thresh = 0.999 * 3 * 255 * args.tile_size * args.tile_size
-        return patch.sum() < thresh
-
     train_transform = Compose(
         [
             LoadImaged(keys=["image"], reader=WSIReader, backend="cucim", dtype=np.uint8, level=1, image_only=True),
@@ -288,7 +283,7 @@ def threshold_filter(patch):
                 patch_size=(args.tile_size, args.tile_size),
                 num_patches=args.tile_count,
                 sort_fn="min",
-                filter_fn=threshold_filter,
+                pad_mode=None,
                 constant_values=255,
             ),
             RandFlipd(keys=["image"], spatial_axis=0, prob=0.5),
@@ -307,7 +302,8 @@ def threshold_filter(patch):
                 keys=["image"],
                 patch_size=(args.tile_size, args.tile_size),
                 sort_fn="min",
-                filter_fn=threshold_filter,
+                threshold=0.999 * 3 * 255 * args.tile_size * args.tile_size,
+                pad_mode=None,
                 constant_values=255,
             ),
             ScaleIntensityRanged(keys=["image"], a_min=np.float32(255), a_max=np.float32(0)),

From ecb1ba11196f4ecfa79cc41cc86a61128b010b39 Mon Sep 17 00:00:00 2001
From: "pre-commit-ci[bot]"
 <66853113+pre-commit-ci[bot]@users.noreply.github.com>
Date: Thu, 2 Jun 2022 00:59:10 +0000
Subject: [PATCH 15/18] [pre-commit.ci] auto fixes from pre-commit.com hooks

for more information, see https://pre-commit.ci
---
 .../panda_mil_train_evaluate_pytorch_gpu.py                   | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/pathology/multiple_instance_learning/panda_mil_train_evaluate_pytorch_gpu.py b/pathology/multiple_instance_learning/panda_mil_train_evaluate_pytorch_gpu.py
index e0ab543e5e..ebe8112d2c 100644
--- a/pathology/multiple_instance_learning/panda_mil_train_evaluate_pytorch_gpu.py
+++ b/pathology/multiple_instance_learning/panda_mil_train_evaluate_pytorch_gpu.py
@@ -35,7 +35,7 @@
 )
 from monai.networks.nets import milmodel
 
-    
+
 def train_epoch(model, loader, optimizer, scaler, epoch, args):
     """One train epoch over the dataset"""
 
@@ -238,7 +238,7 @@ def list_data_collate(batch: collections.abc.Sequence):
         data["image"] = torch.stack([ix["image"] for ix in item], dim=0)
         batch[i] = data
     return default_collate(batch)
-    
+
 def main_worker(gpu, args):
 
     args.gpu = gpu

From a7749734940980026842f043ef3007006a2bfdb3 Mon Sep 17 00:00:00 2001
From: Behrooz <3968947+drbeh@users.noreply.github.com>
Date: Tue, 7 Jun 2022 03:15:27 +0000
Subject: [PATCH 16/18] Update mil pipeline with new grid patch

Signed-off-by: Behrooz <3968947+drbeh@users.noreply.github.com>
---
 .../panda_mil_train_evaluate_pytorch_gpu.py                | 7 +++++--
 1 file changed, 5 insertions(+), 2 deletions(-)

diff --git a/pathology/multiple_instance_learning/panda_mil_train_evaluate_pytorch_gpu.py b/pathology/multiple_instance_learning/panda_mil_train_evaluate_pytorch_gpu.py
index ebe8112d2c..f8d3727d6c 100644
--- a/pathology/multiple_instance_learning/panda_mil_train_evaluate_pytorch_gpu.py
+++ b/pathology/multiple_instance_learning/panda_mil_train_evaluate_pytorch_gpu.py
@@ -239,6 +239,7 @@ def list_data_collate(batch: collections.abc.Sequence):
         batch[i] = data
     return default_collate(batch)
 
+
 def main_worker(gpu, args):
 
     args.gpu = gpu
@@ -282,7 +283,8 @@ def main_worker(gpu, args):
                 keys=["image"],
                 patch_size=(args.tile_size, args.tile_size),
                 num_patches=args.tile_count,
-                sort_fn="min",
+                filter_high_values=True,
+                return_location=False,
                 pad_mode=None,
                 constant_values=255,
             ),
@@ -301,8 +303,9 @@ def main_worker(gpu, args):
             GridPatchd(
                 keys=["image"],
                 patch_size=(args.tile_size, args.tile_size),
-                sort_fn="min",
+                filter_high_values=True,
                 threshold=0.999 * 3 * 255 * args.tile_size * args.tile_size,
+                return_location=False,
                 pad_mode=None,
                 constant_values=255,
             ),

From 34195802d42c255797754099a9104e296e9fa526 Mon Sep 17 00:00:00 2001
From: Behrooz <3968947+drbeh@users.noreply.github.com>
Date: Thu, 9 Jun 2022 00:27:21 +0000
Subject: [PATCH 17/18] Update GridPatch args

Signed-off-by: Behrooz <3968947+drbeh@users.noreply.github.com>
---
 .../panda_mil_train_evaluate_pytorch_gpu.py                  | 5 +----
 1 file changed, 1 insertion(+), 4 deletions(-)

diff --git a/pathology/multiple_instance_learning/panda_mil_train_evaluate_pytorch_gpu.py b/pathology/multiple_instance_learning/panda_mil_train_evaluate_pytorch_gpu.py
index f8d3727d6c..ebe490265f 100644
--- a/pathology/multiple_instance_learning/panda_mil_train_evaluate_pytorch_gpu.py
+++ b/pathology/multiple_instance_learning/panda_mil_train_evaluate_pytorch_gpu.py
@@ -283,8 +283,7 @@ def main_worker(gpu, args):
                 keys=["image"],
                 patch_size=(args.tile_size, args.tile_size),
                 num_patches=args.tile_count,
-                filter_high_values=True,
-                return_location=False,
+                sort_fn="min",
                 pad_mode=None,
                 constant_values=255,
             ),
@@ -303,9 +302,7 @@ def main_worker(gpu, args):
             GridPatchd(
                 keys=["image"],
                 patch_size=(args.tile_size, args.tile_size),
-                filter_high_values=True,
                 threshold=0.999 * 3 * 255 * args.tile_size * args.tile_size,
-                return_location=False,
                 pad_mode=None,
                 constant_values=255,
             ),

From 1a453054a82040c1c7ffbd19f5ca4a21375853a3 Mon Sep 17 00:00:00 2001
From: Behrooz <3968947+drbeh@users.noreply.github.com>
Date: Fri, 8 Jul 2022 15:55:13 +0000
Subject: [PATCH 18/18] Update LabelEncodeIntegerGraded and sort imports

Signed-off-by: Behrooz <3968947+drbeh@users.noreply.github.com>
---
 .../panda_mil_train_evaluate_pytorch_gpu.py   | 47 ++++++++++---------
 1 file changed, 24 insertions(+), 23 deletions(-)

diff --git a/pathology/multiple_instance_learning/panda_mil_train_evaluate_pytorch_gpu.py b/pathology/multiple_instance_learning/panda_mil_train_evaluate_pytorch_gpu.py
index ebe490265f..3d4ea7cb4d 100644
--- a/pathology/multiple_instance_learning/panda_mil_train_evaluate_pytorch_gpu.py
+++ b/pathology/multiple_instance_learning/panda_mil_train_evaluate_pytorch_gpu.py
@@ -1,39 +1,36 @@
-import os
-import time
-import shutil
 import argparse
 import collections.abc
-import gdown
+import os
+import shutil
+import time
 
+import gdown
 import numpy as np
-from sklearn.metrics import cohen_kappa_score
-
 import torch
-import torch.nn as nn
-from torch.cuda.amp import GradScaler, autocast
-
-from torch.utils.tensorboard import SummaryWriter
-from torch.utils.data.distributed import DistributedSampler
-from torch.utils.data.dataloader import default_collate
-
 import torch.distributed as dist
 import torch.multiprocessing as mp
-
+import torch.nn as nn
+from monai.config import KeysCollection
 from monai.data import Dataset, load_decathlon_datalist
 from monai.data.wsi_reader import WSIReader
 from monai.metrics import Cumulative, CumulativeAverage
+from monai.networks.nets import milmodel
 from monai.transforms import (
-    Transform,
     Compose,
+    GridPatchd,
     LoadImaged,
+    MapTransform,
     RandFlipd,
     RandGridPatchd,
-    GridPatchd,
     RandRotate90d,
     ScaleIntensityRanged,
     ToTensord,
 )
-from monai.networks.nets import milmodel
+from sklearn.metrics import cohen_kappa_score
+from torch.cuda.amp import GradScaler, autocast
+from torch.utils.data.dataloader import default_collate
+from torch.utils.data.distributed import DistributedSampler
+from torch.utils.tensorboard import SummaryWriter
 
 
 def train_epoch(model, loader, optimizer, scaler, epoch, args):
@@ -194,7 +191,7 @@ def save_checkpoint(model, epoch, args, filename="model.pt", best_acc=0):
     print("Saving checkpoint", filename)
 
 
-class LabelEncodeIntegerGraded(Transform):
+class LabelEncodeIntegerGraded(MapTransform):
     """
     Convert an integer label to encoded array representation of length num_classes,
     with 1 filled in up to label index, and 0 otherwise. For example for num_classes=5,
@@ -202,14 +199,18 @@ class LabelEncodeIntegerGraded(Transform):
 
     Args:
         num_classes: the number of classes to convert to encoded format.
-        keys: keys of the corresponding items to be transformed
-            Defaults to ``['label']``.
+        keys: keys of the corresponding items to be transformed. Defaults to ``'label'``.
+        allow_missing_keys: don't raise exception if key is missing.
 
     """
 
-    def __init__(self, num_classes, keys=("label",)):
-        super().__init__()
-        self.keys = keys
+    def __init__(
+        self,
+        num_classes: int,
+        keys: KeysCollection = "label",
+        allow_missing_keys: bool = False,
+    ):
+        super().__init__(keys, allow_missing_keys)
         self.num_classes = num_classes
 
     def __call__(self, data):