diff --git a/README.md b/README.md index a6ae46b..d73f9ba 100644 --- a/README.md +++ b/README.md @@ -11,22 +11,30 @@ This repository contains materials and resources for workshops conducted in 2025 ## How to Contribute 1. Fork the repository. -2. Create a new branch for your feature or bug fix: +2. Create a new branch for your tutorial or bug fix: ```bash - git checkout -b my-feature-branch + git checkout -b my-branch ``` 3. Make your changes and commit them with clear messages: ```bash - git commit -m "Add feature X" + git commit -m "Add function ... to simplify tutorial ... content" ``` 4. Push your branch to your forked repository: ```bash - git push origin my-feature-branch + git push origin my-tutorial-branch ``` 5. Open a pull request to the main repository. Please ensure your contributions adhere to the repository's coding standards and include appropriate documentation. +## Building the Book + +To build the book in development, assuming that the working directory is the project's folder, please call: + +```bash +jupyter-book build . +``` + ## Pre-commit Hooks This repository uses pre-commit hooks to ensure code quality and consistency. To set up pre-commit hooks locally, follow these steps: diff --git a/tutorials/brain-disorder-diagnosis/config.py b/tutorials/brain-disorder-diagnosis/config.py index 8392c67..36bde8b 100644 --- a/tutorials/brain-disorder-diagnosis/config.py +++ b/tutorials/brain-disorder-diagnosis/config.py @@ -1,14 +1,12 @@ -import os from yacs.config import CfgNode -DEFAULT_DIR = os.path.join(os.getcwd(), "data") _C = CfgNode() # Dataset configuration _C.DATASET = CfgNode() # Path to the dataset directory -_C.DATASET.PATH = DEFAULT_DIR +_C.DATASET.PATH = "data" # Name of the brain atlas to use # Available options: # - "aal" (AAL) diff --git a/tutorials/brain-disorder-diagnosis/data.py b/tutorials/brain-disorder-diagnosis/data.py index e563eac..b396973 100644 --- a/tutorials/brain-disorder-diagnosis/data.py +++ b/tutorials/brain-disorder-diagnosis/data.py @@ -27,8 +27,8 @@ } ) ], - "vectorize": [bool], - "verbose": [bool], + "vectorize": ["boolean"], + "verbose": ["boolean"], }, prefer_skip_nested_validation=False, ) diff --git a/tutorials/brain-disorder-diagnosis/notebook.ipynb b/tutorials/brain-disorder-diagnosis/notebook.ipynb index 2ea8c10..d17502c 100644 --- a/tutorials/brain-disorder-diagnosis/notebook.ipynb +++ b/tutorials/brain-disorder-diagnosis/notebook.ipynb @@ -1,742 +1,742 @@ { - "nbformat": 4, - "nbformat_minor": 5, - "metadata": { - "kernelspec": { - "display_name": "embc25", - "language": "python", - "name": "python3" - } - }, - "cells": [ - { - "metadata": {}, - "source": [ - "# Brain Disorder Diagnosis\n", - "\n", - "In this tutorial, we demonstrate how to leverage **patient phenotypic information** to reduce **site-specific biases** in functional connectivity data using **domain adaptation**, with the goal of improving **multi-site autism classification**.\n", - "\n", - "This notebook builds on the work of **Kunda et al. (IEEE TMI, 2022)**, which introduced a second-order functional connectivity representation called **Tangent-Pearson**, the tangent embedding of the Pearson correlation matrix. The original work also applied domain adaptation to reduce site dependencies in fMRI-derived features, using **site labels** as the domain information.\n", - "\n", - "We extend this approach by incorporating a **richer set** of phenotypic variables, such as sex, handedness, age, and eye status into the domain adaptation framework. This enables more effective harmonization across data collected from different imaging sites.\n", - "\n", - "---\n", - "\n", - "**Objectives**\n", - "\n", - "1.\t**Load** the ABIDE dataset using different preprocessing pipelines and brain atlases.\n", - "2.\t**Preprocess** phenotypic variables for use in domain adaptation, and obtain class labels (ASD vs CONTROL) and site labels.\n", - "3.\t**Extract** functional connectivity **embedding** from ROI-based time series.\n", - "4.\t**Build** a **training** and **evaluation** pipeline to assess classification performance under various domain adaptation strategies.\n", - "5.\t**Interpret** the learned model by extracting weights for pairwise ROI feature importance and visualizing them using a connectome plot." - ], - "cell_type": "markdown" - }, - { - "metadata": { - "tags": [] - }, - "source": [ - "## Setup\n", - "\n", - "As a starting point, we will install the required packages and load a set of helper functions to assist throughout this tutorial. To keep the output clean and focused on interpretation, we will also suppress warnings.\n", - "\n", - "In addition, several helper scripts are provided to modularize the code and simplify the workflow. These can be inspected directly as `.py` files in the notebook’s current directory. The helper scripts include:\n", - "\n", - "- **`config.py`**: Defines the base configuration settings, which can be customized and overridden using external `.yml` files.\n", - "- **`data.py`**: Provides data loading functions and utilities to automatically download any required datasets.\n", - "- **`parsing.py`**: Contains utilities to compile and summarize evaluation results from the training process.\n", - "- **`preprocess.py`**: Handles phenotype preprocessing, including missing value imputation, categorical variable encoding, and feature extraction from fMRI time series data." - ], - "cell_type": "markdown" - }, - { - "metadata": { - "tags": [ - "hide-input" - ] - }, - "source": [ - "import os\n", - "import site\n", - "import sys\n", - "import warnings\n", - "\n", - "warnings.filterwarnings(\"ignore\")\n", - "os.environ[\"PYTHONWARNINGS\"] = \"ignore\"\n", - "\n", - "if \"google.colab\" in str(get_ipython()):\n", - " sys.path.insert(0, site.getusersitepackages())\n", - " !git clone --single-branch https://github.com/pykale/embc-mmai25.git\n", - " %cp -r /content/embc-mmai25/tutorials/brain-disorder-diagnosis/* /content/\n", - " %rm -r /content/embc-mmai25" - ], - "cell_type": "code", - "outputs": [], - "execution_count": null - }, - { - "metadata": { - "tags": [] - }, - "source": [ - "## Packages\n", - "\n", - "The main packages required for this tutorial are:\n", - "\n", - "- **pykale**: An open-source interdisciplinary machine learning library developed at the University of Sheffield. It focuses on applications in biomedical and scientific domains, providing tools for multimodal learning, domain adaptation, and model interpretability.\n", - "\n", - "- **gdown**: A utility package that simplifies downloading files and folders directly from Google Drive.\n", - "\n", - "- **nilearn**: A Python library for neuroimaging analysis. It offers convenient tools for processing, analyzing, and visualizing functional MRI (fMRI) data.\n", - "\n", - "- **yacs**: A lightweight configuration management library used to store and organize experiment settings in a hierarchical and human-readable format." - ], - "cell_type": "markdown" - }, - { - "metadata": { - "tags": [ - "hide-input" - ] - }, - "source": [ - "!pip install --quiet --user \\\n", - " git+https://github.com/pykale/pykale@main \\\n", - " gdown==5.2.0 nilearn==0.10.4 yacs==0.1.8 \\\n", - " && echo \"pykale, gdown, nilearn, and yacs installed successfully ✅\" \\\n", - " || echo \"Failed to install pykale, gdown, nilearn, and yacs ❌\"" - ], - "cell_type": "code", - "outputs": [ - { - "output_type": "stream", - "name": "stdout", - "text": [ - "pykale, gdown, nilearn, and yacs installed successfully ✅\n" - ] + "nbformat": 4, + "nbformat_minor": 5, + "metadata": { + "kernelspec": { + "display_name": "embc", + "language": "python", + "name": "python3" } - ], - "execution_count": null - }, - { - "metadata": {}, - "source": [ - "## Configuration\n", - "\n", - "To minimize the footprint of the notebook when specifying configurations, we provide a `config.py` file that defines default parameters. These can be customized by supplying a `.yml` configuration file, such as `experiments/base.yml` as an example.\n", - "\n", - "Please refer to these files for detailed instructions on how to customize the experiment settings. \n", - "We provide detailed descriptions of each configurable option in the following sections." - ], - "cell_type": "markdown" }, - { - "metadata": { - "tags": [ - "hide-input" - ] - }, - "source": [ - "from config import get_cfg_defaults\n", - "\n", - "cfg = get_cfg_defaults()\n", - "cfg.merge_from_file(\"experiments/base.yml\")\n", - "cfg.freeze()\n", - "print(cfg)" - ], - "cell_type": "code", - "outputs": [ + "cells": [ { - "output_type": "stream", - "name": "stdout", - "text": [ - "CROSS_VALIDATION:\n", - " NUM_FOLDS: 10\n", - " NUM_REPEATS: 1\n", - " SPLIT: skf\n", - "DATASET:\n", - " ATLAS: hcp-ica\n", - " FC: tangent-pearson\n", - " PATH: /home/zarizky/projects/embc-mmai25/tutorials/brain-disorder-diagnosis/data\n", - "PHENOTYPE:\n", - " STANDARDIZE: site\n", - "RANDOM_STATE: 0\n", - "TRAINER:\n", - " CLASSIFIER: lr\n", - " NONLINEAR: False\n", - " NUM_SEARCH_ITER: 20\n", - " NUM_SOLVER_ITER: 100\n", - " N_JOBS: -1\n", - " PRE_DISPATCH: 2*n_jobs\n", - " REFIT: accuracy\n", - " SCORING: ['accuracy', 'roc_auc']\n", - " SEARCH_STRATEGY: random\n", - " VERBOSE: 0\n" - ] - } - ], - "execution_count": null - }, - { - "metadata": { - "tags": [] - }, - "source": [ - "# Data Loading and Preprocessing\n", - "\n", - "Typically, raw fMRI scans require extensive preprocessing before they can be used in a machine learning pipeline. However, the **ABIDE** dataset provides several preprocessed derivatives, which can be downloaded directly from the [Preprocessed Connectomes Project (PCP)](https://preprocessed-connectomes-project.org/abide/), eliminating the need for manual preprocessing.\n", - "\n", - "Given the long runtime required to extract the functional connectivity embedding, we will omit this step from this notebook and provide pre-computed embeddings through the provided `load_data` function with the associated atlas.\n", - "\n", - "For users interested in computing the time series and functional connectivity embeddings from scratch, assuming preprocessed images are available, please refer to:\n", - "\n", - "- [`NiftiLabelsMasker` (Deterministic / 3D Atlas)](https://nilearn.github.io/stable/modules/generated/nilearn.maskers.NiftiLabelsMasker.html)\n", - "- [`NiftiMapsMasker` (Probabilistic / 4D Atlas)](https://nilearn.github.io/stable/modules/generated/nilearn.maskers.NiftiMapsMasker.html)\n", - "- `extract_functional_connectivity` function implemented in `preprocess.py`.\n", - "\n", - "In this tutorial, we focus on the following preprocessing options:\n", - "\n", - "- **`path`** (or `data_dir`): Directory where the preprocessed dataset is located.\n", - " - *Default:* Current working directory + `/data`\n", - "\n", - "- **`atlas`**: The brain atlas used to extract ROI time series.\n", - " - Available options:\n", - " - `\"aal\"`: AAL Atlas\n", - " - `\"cc200\"`: Cameron Craddock 200\n", - " - `\"cc400\"`: Cameron Craddock 400\n", - " - `\"difumo64\"`: DiFuMo 64\n", - " - `\"dos160\"`: Dosenbach 160\n", - " - `\"hcp-ica\"`: HCP-ICA\n", - " - `\"ho\"`: Harvard-Oxford\n", - " - `\"tt\"`: Talairach-Tournoux \n", - " - *Default:* `\"cc200\"`\n", - "\n", - "- **`fc`**: The functional connectivity measure used to compute pairwise associations between ROIs.\n", - " - Available options:\n", - " - `\"pearson\"`: Pearson correlation\n", - " - `\"partial\"`: Partial correlation\n", - " - `\"tangent\"`: Tangent embedding\n", - " - `\"precision\"`: Precision (inverse covariance)\n", - " - `\"covariance\"`: Covariance\n", - " - `\"tangent-pearson\"`: Tangent-Pearson hybrid connectivity \n", - " - *Default:* `\"tangent-pearson\"`" - ], - "cell_type": "markdown" - }, - { - "metadata": { - "tags": [] - }, - "source": [ - "from data import load_data\n", - "\n", - "fc, phenotypes, rois, coords = load_data(\n", - " cfg.DATASET.PATH, cfg.DATASET.ATLAS, cfg.DATASET.FC\n", - ")" - ], - "cell_type": "code", - "outputs": [ + "metadata": {}, + "source": [ + "# Brain Disorder Diagnosis\n", + "\n", + "In this tutorial, we demonstrate how to leverage **patient phenotypic information** to reduce **site-specific biases** in functional connectivity data using **domain adaptation**, with the goal of improving **multi-site autism classification**.\n", + "\n", + "This notebook builds on the work of **Kunda et al. (IEEE TMI, 2022)**, which introduced a second-order functional connectivity representation called **Tangent-Pearson**, the tangent embedding of the Pearson correlation matrix. The original work also applied domain adaptation to reduce site dependencies in fMRI-derived features, using **site labels** as the domain information.\n", + "\n", + "We extend this approach by incorporating a **richer set** of phenotypic variables, such as sex, handedness, age, and eye status into the domain adaptation framework. This enables more effective harmonization across data collected from different imaging sites.\n", + "\n", + "---\n", + "\n", + "**Objectives**\n", + "\n", + "1.\t**Load** the ABIDE dataset using different preprocessing pipelines and brain atlases.\n", + "2.\t**Preprocess** phenotypic variables for use in domain adaptation, and obtain class labels (ASD vs CONTROL) and site labels.\n", + "3.\t**Extract** functional connectivity **embedding** from ROI-based time series.\n", + "4.\t**Build** a **training** and **evaluation** pipeline to assess classification performance under various domain adaptation strategies.\n", + "5.\t**Interpret** the learned model by extracting weights for pairwise ROI feature importance and visualizing them using a connectome plot." + ], + "cell_type": "markdown" + }, { - "output_type": "stream", - "name": "stdout", - "text": [ - "✔ File found: /home/zarizky/projects/embc-mmai25/tutorials/brain-disorder-diagnosis/data/abide/fc/hcp-ica/tangent-pearson.npy\n", - "✔ File found: /home/zarizky/projects/embc-mmai25/tutorials/brain-disorder-diagnosis/data/abide/phenotypes.csv\n", - "✔ Atlas folder found: /home/zarizky/projects/embc-mmai25/tutorials/brain-disorder-diagnosis/data/atlas/deterministic/hcp-ica\n" - ] - } - ], - "execution_count": null - }, - { - "metadata": { - "tags": [] - }, - "source": [ - "## Phenotype Preprocessing\n", - "\n", - "The phenotypic information in the dataset contains several missing values. We impute and encode these variables to make them suitable for modeling. The `preprocess_phenotypic_data` function handles this functionality for us.\n", - "\n", - "### Categorical Variables\n", - "\n", - "The following categorical phenotypes are used and will be **one-hot encoded**:\n", - "\n", - "- `SITE_ID`\n", - "- `SEX`\n", - "- `HANDEDNESS_CATEGORY`\n", - "- `EYE_STATUS_AT_SCAN`\n", - "\n", - "### Continuous Variables\n", - "\n", - "The following continuous phenotypes will optionally be **standardized**:\n", - "\n", - "- `AGE_AT_SCAN`\n", - "- `FIQ`\n", - "\n", - "Standardization options for continuous phenotypes (`standardize` argument):\n", - "\n", - "- `\"all\"` or `True`: Standardize across all subjects.\n", - "- `\"site\"`: Standardize within each site.\n", - "- `False`: No standardization.\n", - "\n", - "### Handling Missing Values\n", - "\n", - "- `HANDEDNESS_CATEGORY`: Missing values are assumed to correspond to `right-handed` subjects.\n", - "- `FIQ`: Missing values are imputed with a default score of `100`.\n", - "\n", - "### Label Encoding\n", - "\n", - "The diagnostic label `DX_GROUP` is used to assign the target class:\n", - "\n", - "- `CONTROL` → `0`\n", - "- `ASD` → `1`" - ], - "cell_type": "markdown" - }, - { - "metadata": { - "tags": [] - }, - "source": [ - "from preprocess import preprocess_phenotypic_data\n", - "\n", - "labels, sites, phenotypes = preprocess_phenotypic_data(\n", - " phenotypes, cfg.PHENOTYPE.STANDARDIZE\n", - ")" - ], - "cell_type": "code", - "outputs": [], - "execution_count": null - }, - { - "metadata": { - "tags": [] - }, - "source": [ - "# Modeling\n", - "\n", - "We define and train machine learning models for classifying autism spectrum disorder (ASD) using functional connectivity features.\n", - "\n", - "We explore different configurations including a baseline model, domain adaptation using site information, and an extended approach that incorporates additional phenotypic variables.\n", - "\n", - "Each model is evaluated using cross-validation, and we analyze the impact of domain adaptation on classification performance." - ], - "cell_type": "markdown" - }, - { - "metadata": { - "tags": [] - }, - "source": [ - "### Cross-Validation Split\n", - "\n", - "To evaluate model performance reliably, we define a cross-validation (CV) strategy. By default, we use **Repeated Stratified K-Fold**, which preserves class distribution across folds and supports repeated trials for more stable estimates.\n", - "\n", - "Alternatively, we can also use **Leave-P-Groups-Out (LPGO)** cross-validation. This strategy is particularly useful in multi-site studies, as it ensures that data from the same group (e.g., imaging site) are not shared between training and test sets, enabling more realistic generalization assessment under domain shift.\n", - "\n", - "In this tutorial, we specify the following arguments:\n", - "\n", - "- **`split`**: Defines the cross-validation strategy.\n", - " - Available options: \n", - " - `\"skf\"`: Stratified K-Fold to maintain label balance in each fold.\n", - " - `\"lpgo\"`: Leave-P-Groups-Out to evaluate generalization across sites by holding out entire groups (e.g., imaging sites).\n", - " - *Default:* `\"skf\"`\n", - "\n", - "- **`num_folds`**: The number of folds for Stratified K-Fold or the number of groups to leave out in LPGO.\n", - " - *Default:* `10`\n", - "\n", - "- **`num_repeats`**: The number of times the k-fold procedure is repeated to obtain more stable estimates (ignored when using LPGO).\n", - " - *Default:* `5`\n", - "\n", - "- **`random_state`**: Seed for random number generators for reproducibility.\n", - " - *Default:* `None`" - ], - "cell_type": "markdown" - }, - { - "metadata": { - "tags": [] - }, - "source": [ - "from sklearn.model_selection import LeavePGroupsOut, RepeatedStratifiedKFold\n", - "\n", - "# Define the default cross-validation strategy:\n", - "# Repeated stratified k-fold maintains class distribution across folds and supports multiple repetitions\n", - "cv = RepeatedStratifiedKFold(\n", - " # Number of stratified folds\n", - " n_splits=cfg.CROSS_VALIDATION.NUM_FOLDS,\n", - " # Number of repeat rounds\n", - " n_repeats=cfg.CROSS_VALIDATION.NUM_REPEATS,\n", - " # Ensures reproducibility, intentionally set to the seed to have the same splits across runs\n", - " random_state=cfg.RANDOM_STATE,\n", - ")\n", - "\n", - "# Override with leave-p-proups-out if specified\n", - "# This strategy holds out `p` unique groups (e.g., sites) per fold, enabling group-level generalization\n", - "if cfg.CROSS_VALIDATION.SPLIT == \"lpgo\":\n", - " # Use group-based CV for domain adaptation or site bias evaluation\n", - " cv = LeavePGroupsOut(cfg.CROSS_VALIDATION.NUM_FOLDS)" - ], - "cell_type": "code", - "outputs": [], - "execution_count": null - }, - { - "metadata": { - "tags": [] - }, - "source": [ - "### Model Definition\n", - "\n", - "We define several model configurations used for classification. Each model shares the same base classifier but differs in how domain adaptation is applied:\n", - "\n", - "- **Baseline**: A standard model trained directly on functional connectivity features without domain adaptation.\n", - "- **Site Only**: A domain-adapted model that uses site labels as the adaptation factor to reduce site-specific bias.\n", - "- **All Phenotypes**: An extended domain-adapted model that incorporates multiple phenotypic variables (e.g., age, sex, handedness) to further reduce inter-site variability.\n", - "\n", - "We also specify the hyperparameter search strategy and other training parameters for each configuration, including:\n", - "\n", - "- **`classifier`**: The base model used for classification.\n", - " - Available options:\n", - " - `\"lda\"`: Linear Discriminant Analysis\n", - " - `\"lr\"`: Logistic Regression\n", - " - `\"linear_svm\"`: Linear Support Vector Machine\n", - " - `\"svm\"`: Kernel Support Vector Machine\n", - " - `\"ridge\"`: Ridge Classifier (L2-regularized linear model)\n", - " - `\"auto\"`: Automatically selects an appropriate model based on data characteristics.\n", - " - *Default:* `\"lr\"`\n", - "\n", - "- **`nonlinear`**: Whether to apply non-linear transformations (non-interpretable).\n", - " - *Type:* `boolean`\n", - " - *Default:* `False`\n", - "\n", - "- **`search_strategy`**: The hyperparameter search method.\n", - " - Available options:\n", - " - `\"random\"`: Randomly search over finite iterations.\n", - " - `\"grid\"`: Search over all possible combinations.\n", - " - *Default:* `\"random\"`\n", - "\n", - "- **`num_search_iterations`**: The number of hyperparameter combinations to evaluate in randomized search.\n", - " - *Default:* `1,000`\n", - "\n", - "- **`num_solver_iterations`**: The maximum number of iterations allowed for solver convergence.\n", - " - *Default:* `1,000,000`\n", - "\n", - "- **`scoring`**: A list of performance metrics used during cross-validation.\n", - " - Available options:\n", - " - `\"accuracy\"`: Accuracy\n", - " - `\"precision\"`: Precision\n", - " - `\"recall\"`: Recall\n", - " - `\"f1\"`: F1-Score\n", - " - `\"roc_auc\"`: Area Under ROC Curve (AUROC)\n", - " - `\"matthews_corrcoef\"`: Matthews Correlation Coefficient (MCC)\n", - " - *Default:* `[\"accuracy\", \"roc_auc\"]`\n", - "\n", - "- **`refit`**: The metric used to refit the best model after hyperparameter tuning.\n", - " - *Default:* `\"accuracy\"`\n", - "\n", - "- **`num_jobs`**: The number of CPU cores used for training and hyperparameter search.\n", - " - Set to `-1` for all CPUs, `-k` for all but `k` CPUs.\n", - " - *Default:* `-1`\n", - "\n", - "- **`pre_dispatch`**: Controls job pre-dispatching for parallel execution.\n", - " - *Default:* `\"2*n_jobs\"`\n", - "\n", - "- **`verbose`**: Controls verbosity of training output.\n", - " - *Default:* `0`\n", - "\n", - "- **`random_state`**: Seed for random number generators for reproducibility.\n", - " - *Default:* `None`" - ], - "cell_type": "markdown" - }, - { - "metadata": { - "tags": [] - }, - "source": [ - "from kale.pipeline.multi_domain_adapter import AutoMIDAClassificationTrainer as Trainer\n", - "from sklearn.base import clone\n", - "\n", - "# Configuration with cv and random_state/seed included\n", - "trainer_cfg = {k.lower(): v for k, v in cfg.TRAINER.items()}\n", - "trainer_cfg = {**trainer_cfg, \"cv\": cv, \"random_state\": cfg.RANDOM_STATE}\n", - "\n", - "# Initialize dictionary for different trainers\n", - "trainers = {}\n", - "\n", - "# Create a baseline trainer without domain adaptation (MIDA disabled)\n", - "trainers[\"baseline\"] = Trainer(use_mida=False, **trainer_cfg)\n", - "\n", - "# Create a trainer with MIDA enabled, using site labels as domain adaptation factors\n", - "trainers[\"site_only\"] = Trainer(use_mida=True, **trainer_cfg)\n", - "\n", - "# Clone the 'site_only' trainer to create 'all_phenotypes' trainer\n", - "# This enables reusing the same training configuration, while modifying only the input domain factors\n", - "trainers[\"all_phenotypes\"] = clone(trainers[\"site_only\"])" - ], - "cell_type": "code", - "outputs": [], - "execution_count": null - }, - { - "metadata": { - "tags": [] - }, - "source": [ - "# Training\n", - "\n", - "We train each model configuration using the previously defined cross-validation strategy. The training process involves fitting the model on functional connectivity features and evaluating its performance using multiple scoring metrics (e.g., accuracy, F1-score, AUROC).\n", - "\n", - "For models with domain adaptation, we pass additional domain factors (such as site or phenotypic variables) to guide the alignment of embedding. Cross-validation is performed to ensure robust performance estimates and to select the best hyperparameter configuration for each model." - ], - "cell_type": "markdown" - }, - { - "metadata": { - "tags": [] - }, - "source": [ - "import pandas as pd\n", - "from tqdm import tqdm\n", - "\n", - "# Define common training arguments for all models: features (X), labels (y), and group info (sites)\n", - "fit_args = {\"x\": fc, \"y\": labels, \"groups\": sites}\n", - "\n", - "cv_results = {}\n", - "for model in (pbar := tqdm(trainers)):\n", - " args = clone(fit_args, safe=False)\n", - " if model == \"site_only\":\n", - " args[\"group_labels\"] = sites\n", - " if model == \"all_phenotypes\":\n", - " args[\"group_labels\"] = phenotypes\n", - "\n", - " pbar.set_description(f\"Fitting {model} model\")\n", - " trainers[model].fit(**args)\n", - " cv_results[model] = pd.DataFrame(trainers[model].cv_results_)" - ], - "cell_type": "code", - "outputs": [ + "metadata": { + "tags": [] + }, + "source": [ + "# Setup\n", + "\n", + "As a starting point, we will install the required packages and load a set of helper functions to assist throughout this tutorial. To keep the output clean and focused on interpretation, we will also suppress warnings.\n", + "\n", + "In addition, several helper scripts are provided to modularize the code and simplify the workflow. These can be inspected directly as `.py` files in the notebook\u2019s current directory. The helper scripts include:\n", + "\n", + "- **`config.py`**: Defines the base configuration settings, which can be customized and overridden using external `.yml` files.\n", + "- **`data.py`**: Provides data loading functions and utilities to automatically download any required datasets.\n", + "- **`parsing.py`**: Contains utilities to compile and summarize evaluation results from the training process.\n", + "- **`preprocess.py`**: Handles phenotype preprocessing, including missing value imputation, categorical variable encoding, and feature extraction from fMRI time series data." + ], + "cell_type": "markdown" + }, { - "output_type": "stream", - "name": "stderr", - "text": [ - "Fitting all_phenotypes model: 100%|██████████| 3/3 [00:21<00:00, 7.23s/it]\n" - ] - } - ], - "execution_count": null - }, - { - "metadata": { - "tags": [] - }, - "source": [ - "# Evaluation\n", - "\n", - "We evaluate and compare the performance of different model configurations using cross-validation results. We aggregate the top-performing scores for each model based on a specified evaluation metric (e.g., accuracy), allowing us to assess the effectiveness of domain adaptation strategies.\n", - "\n", - "By comparing models with and without domain adaptation, we can determine the impact of incorporating site and phenotypic information on multi-site autism classification performance. This analysis helps identify which configurations generalize best across heterogeneous imaging sites." - ], - "cell_type": "markdown" - }, - { - "metadata": { - "tags": [] - }, - "source": [ - "from parsing import compile_results\n", - "\n", - "# Compile the cross-validation results into a summary table,\n", - "# sorting by the model with the highest test accuracy across CV folds\n", - "compiled_results = compile_results(cv_results, \"accuracy\")\n", - "\n", - "# Display the compiled results DataFrame (models as rows, metrics as formatted strings)\n", - "display(compiled_results)" - ], - "cell_type": "code", - "outputs": [ + "metadata": { + "tags": [ + "hide-input" + ] + }, + "source": [ + "import os\n", + "import site\n", + "import sys\n", + "import warnings\n", + "\n", + "warnings.filterwarnings(\"ignore\")\n", + "os.environ[\"PYTHONWARNINGS\"] = \"ignore\"\n", + "\n", + "if \"google.colab\" in str(get_ipython()):\n", + " sys.path.insert(0, site.getusersitepackages())\n", + " !git clone --single-branch https://github.com/pykale/embc-mmai25.git\n", + " %cp -r /content/embc-mmai25/tutorials/brain-disorder-diagnosis/* /content/\n", + " %rm -r /content/embc-mmai25" + ], + "cell_type": "code", + "outputs": [], + "execution_count": null + }, { - "output_type": "display_data", - "data": { - "text/html": [ - "
\n", - "\n", - "\n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - "
AccuracyAUROC
Model
Baseline0.6629 ± 0.05230.7105 ± 0.0556
Site Only0.6609 ± 0.05090.7127 ± 0.0596
All Phenotypes0.6474 ± 0.05970.7057 ± 0.0514
\n", - "
" + "metadata": { + "tags": [] + }, + "source": [ + "## Packages\n", + "\n", + "The main packages required for this tutorial are:\n", + "\n", + "- **pykale**: An open-source interdisciplinary machine learning library developed at the University of Sheffield. It focuses on applications in biomedical and scientific domains, providing tools for multimodal learning, domain adaptation, and model interpretability.\n", + "\n", + "- **gdown**: A utility package that simplifies downloading files and folders directly from Google Drive.\n", + "\n", + "- **nilearn**: A Python library for neuroimaging analysis. It offers convenient tools for processing, analyzing, and visualizing functional MRI (fMRI) data.\n", + "\n", + "- **yacs**: A lightweight configuration management library used to store and organize experiment settings in a hierarchical and human-readable format." ], - "text/plain": [ - " Accuracy AUROC\n", - "Model \n", - "Baseline 0.6629 ± 0.0523 0.7105 ± 0.0556\n", - "Site Only 0.6609 ± 0.0509 0.7127 ± 0.0596\n", - "All Phenotypes 0.6474 ± 0.0597 0.7057 ± 0.0514" - ] - }, - "metadata": {} - } - ], - "execution_count": null - }, - { - "metadata": { - "tags": [] - }, - "source": [ - "# Interpretation" - ], - "cell_type": "markdown" - }, - { - "metadata": { - "tags": [] - }, - "source": [ - "We interpret the trained models by analyzing the learned weights associated with functional connectivity features. Specifically, we extract the top-weighted ROI pairs that contributed most to the classification decision.\n", - "\n", - "These weights are visualized as a **connectome plot**, allowing us to examine which brain region interactions are most informative for distinguishing individuals with autism from controls. This not only enhances the interpretability of the model but also provides potential insights into neurobiological patterns relevant to autism." - ], - "cell_type": "markdown" - }, - { - "metadata": { - "tags": [] - }, - "source": [ - "import numpy as np\n", - "from kale.interpret.visualize import visualize_connectome\n", - "\n", - "# Fetch coefficients to visualize feature importance\n", - "coef = trainers[\"site_only\"].coef_.ravel()\n", - "# check if coef != features, assumes augmented features with phenotypes/sites\n", - "if coef.shape[0] != fc.shape[1]:\n", - " coef, _ = np.split(coef, [fc.shape[1]])\n", - "\n", - "# Visualize the coefficients as a connectome plot\n", - "proj = visualize_connectome(\n", - " trainers[\"baseline\"].coef_.ravel(),\n", - " rois,\n", - " coords,\n", - " 0.015, # Take top 1.5% of connections\n", - " legend_params={\n", - " \"bbox_to_anchor\": (2.75, -0.1), # Align legend outside the plot\n", - " \"ncol\": 2,\n", - " },\n", - ")\n", - "\n", - "# Display the resulting connectome plot\n", - "display(proj)" - ], - "cell_type": "code", - "outputs": [ + "cell_type": "markdown" + }, + { + "metadata": { + "tags": [ + "hide-input" + ] + }, + "source": [ + "!pip install --quiet --user \\\n", + " git+https://github.com/pykale/pykale@main \\\n", + " gdown==5.2.0 nilearn==0.10.4 yacs==0.1.8 \\\n", + " && echo \"pykale, gdown, nilearn, and yacs installed successfully \u2705\" \\\n", + " || echo \"Failed to install pykale, gdown, nilearn, and yacs \u274c\"" + ], + "cell_type": "code", + "outputs": [ + { + "output_type": "stream", + "name": "stdout", + "text": [ + "pykale, gdown, nilearn, and yacs installed successfully \u2705\n" + ] + } + ], + "execution_count": null + }, + { + "metadata": {}, + "source": [ + "## Configuration\n", + "\n", + "To minimize the footprint of the notebook when specifying configurations, we provide a `config.py` file that defines default parameters. These can be customized by supplying a `.yml` configuration file, such as `experiments/base.yml` as an example.\n", + "\n", + "Please refer to these files for detailed instructions on how to customize the experiment settings. \n", + "We provide detailed descriptions of each configurable option in the following sections." + ], + "cell_type": "markdown" + }, + { + "metadata": { + "tags": [ + "hide-input" + ] + }, + "source": [ + "from config import get_cfg_defaults\n", + "\n", + "cfg = get_cfg_defaults()\n", + "cfg.merge_from_file(\"experiments/base.yml\")\n", + "cfg.freeze()\n", + "print(cfg)" + ], + "cell_type": "code", + "outputs": [ + { + "output_type": "stream", + "name": "stdout", + "text": [ + "CROSS_VALIDATION:\n", + " NUM_FOLDS: 10\n", + " NUM_REPEATS: 1\n", + " SPLIT: skf\n", + "DATASET:\n", + " ATLAS: hcp-ica\n", + " FC: tangent-pearson\n", + " PATH: data\n", + "PHENOTYPE:\n", + " STANDARDIZE: site\n", + "RANDOM_STATE: 0\n", + "TRAINER:\n", + " CLASSIFIER: lr\n", + " NONLINEAR: False\n", + " NUM_SEARCH_ITER: 20\n", + " NUM_SOLVER_ITER: 100\n", + " N_JOBS: -1\n", + " PRE_DISPATCH: 2*n_jobs\n", + " REFIT: accuracy\n", + " SCORING: ['accuracy', 'roc_auc']\n", + " SEARCH_STRATEGY: random\n", + " VERBOSE: 0\n" + ] + } + ], + "execution_count": null + }, + { + "metadata": { + "tags": [] + }, + "source": [ + "# Data Loading and Preprocessing\n", + "\n", + "Typically, raw fMRI scans require extensive preprocessing before they can be used in a machine learning pipeline. However, the **ABIDE** dataset provides several preprocessed derivatives, which can be downloaded directly from the [Preprocessed Connectomes Project (PCP)](https://preprocessed-connectomes-project.org/abide/), eliminating the need for manual preprocessing.\n", + "\n", + "Given the long runtime required to extract the functional connectivity embedding, we will omit this step from this notebook and provide pre-computed embeddings through the provided `load_data` function with the associated atlas.\n", + "\n", + "For users interested in computing the time series and functional connectivity embeddings from scratch, assuming preprocessed images are available, please refer to:\n", + "\n", + "- [`NiftiLabelsMasker` (Deterministic / 3D Atlas)](https://nilearn.github.io/stable/modules/generated/nilearn.maskers.NiftiLabelsMasker.html)\n", + "- [`NiftiMapsMasker` (Probabilistic / 4D Atlas)](https://nilearn.github.io/stable/modules/generated/nilearn.maskers.NiftiMapsMasker.html)\n", + "- `extract_functional_connectivity` function implemented in `preprocess.py`.\n", + "\n", + "In this tutorial, we focus on the following preprocessing options:\n", + "\n", + "- **`path`** (or `data_dir`): Directory where the preprocessed dataset is located.\n", + " - *Default:* Current working directory + `/data`\n", + "\n", + "- **`atlas`**: The brain atlas used to extract ROI time series.\n", + " - Available options:\n", + " - `\"aal\"`: AAL Atlas\n", + " - `\"cc200\"`: Cameron Craddock 200\n", + " - `\"cc400\"`: Cameron Craddock 400\n", + " - `\"difumo64\"`: DiFuMo 64\n", + " - `\"dos160\"`: Dosenbach 160\n", + " - `\"hcp-ica\"`: HCP-ICA\n", + " - `\"ho\"`: Harvard-Oxford\n", + " - `\"tt\"`: Talairach-Tournoux \n", + " - *Default:* `\"cc200\"`\n", + "\n", + "- **`fc`**: The functional connectivity measure used to compute pairwise associations between ROIs.\n", + " - Available options:\n", + " - `\"pearson\"`: Pearson correlation\n", + " - `\"partial\"`: Partial correlation\n", + " - `\"tangent\"`: Tangent embedding\n", + " - `\"precision\"`: Precision (inverse covariance)\n", + " - `\"covariance\"`: Covariance\n", + " - `\"tangent-pearson\"`: Tangent-Pearson hybrid connectivity \n", + " - *Default:* `\"tangent-pearson\"`" + ], + "cell_type": "markdown" + }, + { + "metadata": { + "tags": [] + }, + "source": [ + "from data import load_data\n", + "\n", + "fc, phenotypes, rois, coords = load_data(\n", + " cfg.DATASET.PATH, cfg.DATASET.ATLAS, cfg.DATASET.FC\n", + ")" + ], + "cell_type": "code", + "outputs": [ + { + "output_type": "stream", + "name": "stdout", + "text": [ + "\u2714 File found: data/abide/fc/hcp-ica/tangent-pearson.npy\n", + "\u2714 File found: data/abide/phenotypes.csv\n", + "\u2714 Atlas folder found: data/atlas/deterministic/hcp-ica\n" + ] + } + ], + "execution_count": null + }, + { + "metadata": { + "tags": [] + }, + "source": [ + "## Phenotype Preprocessing\n", + "\n", + "The phenotypic information in the dataset contains several missing values. We impute and encode these variables to make them suitable for modeling. The `preprocess_phenotypic_data` function handles this functionality for us.\n", + "\n", + "### Categorical Variables\n", + "\n", + "The following categorical phenotypes are used and will be **one-hot encoded**:\n", + "\n", + "- `SITE_ID`\n", + "- `SEX`\n", + "- `HANDEDNESS_CATEGORY`\n", + "- `EYE_STATUS_AT_SCAN`\n", + "\n", + "### Continuous Variables\n", + "\n", + "The following continuous phenotypes will optionally be **standardized**:\n", + "\n", + "- `AGE_AT_SCAN`\n", + "- `FIQ`\n", + "\n", + "Standardization options for continuous phenotypes (`standardize` argument):\n", + "\n", + "- `\"all\"` or `True`: Standardize across all subjects.\n", + "- `\"site\"`: Standardize within each site.\n", + "- `False`: No standardization.\n", + "\n", + "### Handling Missing Values\n", + "\n", + "- `HANDEDNESS_CATEGORY`: Missing values are assumed to correspond to `right-handed` subjects.\n", + "- `FIQ`: Missing values are imputed with a default score of `100`.\n", + "\n", + "### Label Encoding\n", + "\n", + "The diagnostic label `DX_GROUP` is used to assign the target class:\n", + "\n", + "- `CONTROL` \u2192 `0`\n", + "- `ASD` \u2192 `1`" + ], + "cell_type": "markdown" + }, + { + "metadata": { + "tags": [] + }, + "source": [ + "from preprocess import preprocess_phenotypic_data\n", + "\n", + "labels, sites, phenotypes = preprocess_phenotypic_data(\n", + " phenotypes, cfg.PHENOTYPE.STANDARDIZE\n", + ")" + ], + "cell_type": "code", + "outputs": [], + "execution_count": null + }, + { + "metadata": { + "tags": [] + }, + "source": [ + "# Modeling\n", + "\n", + "We define and train machine learning models for classifying autism spectrum disorder (ASD) using functional connectivity features.\n", + "\n", + "We explore different configurations including a baseline model, domain adaptation using site information, and an extended approach that incorporates additional phenotypic variables.\n", + "\n", + "Each model is evaluated using cross-validation, and we analyze the impact of domain adaptation on classification performance." + ], + "cell_type": "markdown" + }, + { + "metadata": { + "tags": [] + }, + "source": [ + "### Cross-Validation Split\n", + "\n", + "To evaluate model performance reliably, we define a cross-validation (CV) strategy. By default, we use **Repeated Stratified K-Fold**, which preserves class distribution across folds and supports repeated trials for more stable estimates.\n", + "\n", + "Alternatively, we can also use **Leave-P-Groups-Out (LPGO)** cross-validation. This strategy is particularly useful in multi-site studies, as it ensures that data from the same group (e.g., imaging site) are not shared between training and test sets, enabling more realistic generalization assessment under domain shift.\n", + "\n", + "In this tutorial, we specify the following arguments:\n", + "\n", + "- **`split`**: Defines the cross-validation strategy.\n", + " - Available options: \n", + " - `\"skf\"`: Stratified K-Fold to maintain label balance in each fold.\n", + " - `\"lpgo\"`: Leave-P-Groups-Out to evaluate generalization across sites by holding out entire groups (e.g., imaging sites).\n", + " - *Default:* `\"skf\"`\n", + "\n", + "- **`num_folds`**: The number of folds for Stratified K-Fold or the number of groups to leave out in LPGO.\n", + " - *Default:* `10`\n", + "\n", + "- **`num_repeats`**: The number of times the k-fold procedure is repeated to obtain more stable estimates (ignored when using LPGO).\n", + " - *Default:* `5`\n", + "\n", + "- **`random_state`**: Seed for random number generators for reproducibility.\n", + " - *Default:* `None`" + ], + "cell_type": "markdown" + }, + { + "metadata": { + "tags": [] + }, + "source": [ + "from sklearn.model_selection import LeavePGroupsOut, RepeatedStratifiedKFold\n", + "\n", + "# Define the default cross-validation strategy:\n", + "# Repeated stratified k-fold maintains class distribution across folds and supports multiple repetitions\n", + "cv = RepeatedStratifiedKFold(\n", + " # Number of stratified folds\n", + " n_splits=cfg.CROSS_VALIDATION.NUM_FOLDS,\n", + " # Number of repeat rounds\n", + " n_repeats=cfg.CROSS_VALIDATION.NUM_REPEATS,\n", + " # Ensures reproducibility, intentionally set to the seed to have the same splits across runs\n", + " random_state=cfg.RANDOM_STATE,\n", + ")\n", + "\n", + "# Override with leave-p-proups-out if specified\n", + "# This strategy holds out `p` unique groups (e.g., sites) per fold, enabling group-level generalization\n", + "if cfg.CROSS_VALIDATION.SPLIT == \"lpgo\":\n", + " # Use group-based CV for domain adaptation or site bias evaluation\n", + " cv = LeavePGroupsOut(cfg.CROSS_VALIDATION.NUM_FOLDS)" + ], + "cell_type": "code", + "outputs": [], + "execution_count": null + }, + { + "metadata": { + "tags": [] + }, + "source": [ + "### Model Definition\n", + "\n", + "We define several model configurations used for classification. Each model shares the same base classifier but differs in how domain adaptation is applied:\n", + "\n", + "- **Baseline**: A standard model trained directly on functional connectivity features without domain adaptation.\n", + "- **Site Only**: A domain-adapted model that uses site labels as the adaptation factor to reduce site-specific bias.\n", + "- **All Phenotypes**: An extended domain-adapted model that incorporates multiple phenotypic variables (e.g., age, sex, handedness) to further reduce inter-site variability.\n", + "\n", + "We also specify the hyperparameter search strategy and other training parameters for each configuration, including:\n", + "\n", + "- **`classifier`**: The base model used for classification.\n", + " - Available options:\n", + " - `\"lda\"`: Linear Discriminant Analysis\n", + " - `\"lr\"`: Logistic Regression\n", + " - `\"linear_svm\"`: Linear Support Vector Machine\n", + " - `\"svm\"`: Kernel Support Vector Machine\n", + " - `\"ridge\"`: Ridge Classifier (L2-regularized linear model)\n", + " - `\"auto\"`: Automatically selects an appropriate model based on data characteristics.\n", + " - *Default:* `\"lr\"`\n", + "\n", + "- **`nonlinear`**: Whether to apply non-linear transformations (non-interpretable).\n", + " - *Type:* `boolean`\n", + " - *Default:* `False`\n", + "\n", + "- **`search_strategy`**: The hyperparameter search method.\n", + " - Available options:\n", + " - `\"random\"`: Randomly search over finite iterations.\n", + " - `\"grid\"`: Search over all possible combinations.\n", + " - *Default:* `\"random\"`\n", + "\n", + "- **`num_search_iterations`**: The number of hyperparameter combinations to evaluate in randomized search.\n", + " - *Default:* `1,000`\n", + "\n", + "- **`num_solver_iterations`**: The maximum number of iterations allowed for solver convergence.\n", + " - *Default:* `1,000,000`\n", + "\n", + "- **`scoring`**: A list of performance metrics used during cross-validation.\n", + " - Available options:\n", + " - `\"accuracy\"`: Accuracy\n", + " - `\"precision\"`: Precision\n", + " - `\"recall\"`: Recall\n", + " - `\"f1\"`: F1-Score\n", + " - `\"roc_auc\"`: Area Under ROC Curve (AUROC)\n", + " - `\"matthews_corrcoef\"`: Matthews Correlation Coefficient (MCC)\n", + " - *Default:* `[\"accuracy\", \"roc_auc\"]`\n", + "\n", + "- **`refit`**: The metric used to refit the best model after hyperparameter tuning.\n", + " - *Default:* `\"accuracy\"`\n", + "\n", + "- **`num_jobs`**: The number of CPU cores used for training and hyperparameter search.\n", + " - Set to `-1` for all CPUs, `-k` for all but `k` CPUs.\n", + " - *Default:* `-1`\n", + "\n", + "- **`pre_dispatch`**: Controls job pre-dispatching for parallel execution.\n", + " - *Default:* `\"2*n_jobs\"`\n", + "\n", + "- **`verbose`**: Controls verbosity of training output.\n", + " - *Default:* `0`\n", + "\n", + "- **`random_state`**: Seed for random number generators for reproducibility.\n", + " - *Default:* `None`" + ], + "cell_type": "markdown" + }, + { + "metadata": { + "tags": [] + }, + "source": [ + "from sklearn.base import clone\n", + "from kale.pipeline.multi_domain_adapter import AutoMIDAClassificationTrainer as Trainer\n", + "\n", + "# Configuration with cv and random_state/seed included\n", + "trainer_cfg = {k.lower(): v for k, v in cfg.TRAINER.items()}\n", + "trainer_cfg = {**trainer_cfg, \"cv\": cv, \"random_state\": cfg.RANDOM_STATE}\n", + "\n", + "# Initialize dictionary for different trainers\n", + "trainers = {}\n", + "\n", + "# Create a baseline trainer without domain adaptation (MIDA disabled)\n", + "trainers[\"baseline\"] = Trainer(use_mida=False, **trainer_cfg)\n", + "\n", + "# Create a trainer with MIDA enabled, using site labels as domain adaptation factors\n", + "trainers[\"site_only\"] = Trainer(use_mida=True, **trainer_cfg)\n", + "\n", + "# Clone the 'site_only' trainer to create 'all_phenotypes' trainer\n", + "# This enables reusing the same training configuration, while modifying only the input domain factors\n", + "trainers[\"all_phenotypes\"] = clone(trainers[\"site_only\"])" + ], + "cell_type": "code", + "outputs": [], + "execution_count": null + }, { - "output_type": "display_data", - "data": { - "text/plain": [ - "" - ] - }, - "metadata": {} + "metadata": { + "tags": [] + }, + "source": [ + "# Training\n", + "\n", + "We train each model configuration using the previously defined cross-validation strategy. The training process involves fitting the model on functional connectivity features and evaluating its performance using multiple scoring metrics (e.g., accuracy, F1-score, AUROC).\n", + "\n", + "For models with domain adaptation, we pass additional domain factors (such as site or phenotypic variables) to guide the alignment of embedding. Cross-validation is performed to ensure robust performance estimates and to select the best hyperparameter configuration for each model." + ], + "cell_type": "markdown" + }, + { + "metadata": { + "tags": [] + }, + "source": [ + "import pandas as pd\n", + "from tqdm import tqdm\n", + "\n", + "# Define common training arguments for all models: features (X), labels (y), and group info (sites)\n", + "fit_args = {\"x\": fc, \"y\": labels, \"groups\": sites}\n", + "\n", + "cv_results = {}\n", + "for model in (pbar := tqdm(trainers)):\n", + " args = clone(fit_args, safe=False)\n", + " if model == \"site_only\":\n", + " args[\"group_labels\"] = sites\n", + " if model == \"all_phenotypes\":\n", + " args[\"group_labels\"] = phenotypes\n", + "\n", + " pbar.set_description(f\"Fitting {model} model\")\n", + " trainers[model].fit(**args)\n", + " cv_results[model] = pd.DataFrame(trainers[model].cv_results_)" + ], + "cell_type": "code", + "outputs": [ + { + "output_type": "stream", + "name": "stderr", + "text": [ + "Fitting all_phenotypes model: 100%|\u2588\u2588\u2588\u2588\u2588\u2588\u2588\u2588\u2588\u2588| 3/3 [00:24<00:00, 8.19s/it]\n" + ] + } + ], + "execution_count": null + }, + { + "metadata": { + "tags": [] + }, + "source": [ + "# Evaluation\n", + "\n", + "We evaluate and compare the performance of different model configurations using cross-validation results. We aggregate the top-performing scores for each model based on a specified evaluation metric (e.g., accuracy), allowing us to assess the effectiveness of domain adaptation strategies.\n", + "\n", + "By comparing models with and without domain adaptation, we can determine the impact of incorporating site and phenotypic information on multi-site autism classification performance. This analysis helps identify which configurations generalize best across heterogeneous imaging sites." + ], + "cell_type": "markdown" + }, + { + "metadata": { + "tags": [] + }, + "source": [ + "from parsing import compile_results\n", + "\n", + "# Compile the cross-validation results into a summary table,\n", + "# sorting by the model with the highest test accuracy across CV folds\n", + "compiled_results = compile_results(cv_results, \"accuracy\")\n", + "\n", + "# Display the compiled results DataFrame (models as rows, metrics as formatted strings)\n", + "display(compiled_results)" + ], + "cell_type": "code", + "outputs": [ + { + "output_type": "display_data", + "data": { + "text/html": [ + "
\n", + "\n", + "\n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + "
AccuracyAUROC
Model
Baseline0.6629 \u00b1 0.05230.7105 \u00b1 0.0556
Site Only0.6609 \u00b1 0.05090.7127 \u00b1 0.0596
All Phenotypes0.6474 \u00b1 0.05970.7057 \u00b1 0.0514
\n", + "
" + ], + "text/plain": [ + " Accuracy AUROC\n", + "Model \n", + "Baseline 0.6629 \u00b1 0.0523 0.7105 \u00b1 0.0556\n", + "Site Only 0.6609 \u00b1 0.0509 0.7127 \u00b1 0.0596\n", + "All Phenotypes 0.6474 \u00b1 0.0597 0.7057 \u00b1 0.0514" + ] + }, + "metadata": {} + } + ], + "execution_count": null }, { - "output_type": "display_data", - "data": { - "image/png": "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", - "text/plain": [ - "
" - ] - }, - "metadata": {} + "metadata": { + "tags": [] + }, + "source": [ + "# Interpretation" + ], + "cell_type": "markdown" + }, + { + "metadata": { + "tags": [] + }, + "source": [ + "We interpret the trained models by analyzing the learned weights associated with functional connectivity features. Specifically, we extract the top-weighted ROI pairs that contributed most to the classification decision.\n", + "\n", + "These weights are visualized as a **connectome plot**, allowing us to examine which brain region interactions are most informative for distinguishing individuals with autism from controls. This not only enhances the interpretability of the model but also provides potential insights into neurobiological patterns relevant to autism." + ], + "cell_type": "markdown" + }, + { + "metadata": { + "tags": [] + }, + "source": [ + "import numpy as np\n", + "from kale.interpret.visualize import visualize_connectome\n", + "\n", + "# Fetch coefficients to visualize feature importance\n", + "coef = trainers[\"site_only\"].coef_.ravel()\n", + "# check if coef != features, assumes augmented features with phenotypes/sites\n", + "if coef.shape[0] != fc.shape[1]:\n", + " coef, _ = np.split(coef, [fc.shape[1]])\n", + "\n", + "# Visualize the coefficients as a connectome plot\n", + "proj = visualize_connectome(\n", + " trainers[\"baseline\"].coef_.ravel(),\n", + " rois,\n", + " coords,\n", + " 0.015, # Take top 1.5% of connections\n", + " legend_params={\n", + " \"bbox_to_anchor\": (2.75, -0.1), # Align legend outside the plot\n", + " \"ncol\": 2,\n", + " },\n", + ")\n", + "\n", + "# Display the resulting connectome plot\n", + "display(proj)" + ], + "cell_type": "code", + "outputs": [ + { + "output_type": "display_data", + "data": { + "text/plain": [ + "" + ] + }, + "metadata": {} + }, + { + "output_type": "display_data", + "data": { + "image/png": "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", + "text/plain": [ + "
" + ] + }, + "metadata": {} + } + ], + "execution_count": null + }, + { + "metadata": { + "tags": [] + }, + "source": [ + "### Interpretation of Discriminative Connectivity Patterns\n", + "\n", + "This plot shows the **most discriminative ROI connections** for classifying ASD vs Control subjects.\n", + "- **Red edges** indicate connections **stronger in ASD**.\n", + "- **Blue edges** indicate connections **stronger in Control**.\n", + "- Color intensity reflects the **magnitude of contribution** to the model\u2019s decision.\n", + "\n", + "---\n", + "\n", + "**Key Patterns**:\n", + "\n", + "- **Default Mode Network (DMN)**:\n", + " - *DefaultMode.MPFC*, *DefaultMode.PCC*, *DefaultMode.LP (L/R)*\n", + " - Core hubs of the DMN, associated with **self-referential processing**, **social cognition**, and often disrupted in ASD.\n", + "\n", + "- **Fronto-Parietal Network**:\n", + " - *FrontoParietal.PPC (L)*\n", + " - Involved in **executive function** and **cognitive flexibility**, domains typically impaired in ASD.\n", + "\n", + "- **Dorsal Attention Network**:\n", + " - *DorsalAttention.IPS (L)*\n", + " - Associated with **goal-directed attention**, potentially altered in ASD subjects.\n", + "\n", + "- **Salience Network**:\n", + " - *Salience.SMG (R)*\n", + " - Plays a role in **interoception** and **social-emotional processing**, relevant for ASD symptoms.\n", + "\n", + "- **Language Network**:\n", + " - *Language.pSTG (R)*\n", + " - Critical for **language comprehension** and **social communication**, often affected in ASD.\n", + "\n", + "- **Sensorimotor and Cerebellar Regions**:\n", + " - *SensoriMotor.Lateral (L)*, *Cerebellar.Posterior*\n", + " - Linked to **motor coordination** and **sensorimotor integration**, commonly atypical in ASD.\n", + "\n", + "The interpretability analysis of the trained model highlights that **functional connectivity alterations across DMN, attention, salience, language, and sensorimotor systems** are key discriminative factors for distinguishing **ASD** from **Control** subjects." + ], + "cell_type": "markdown" } - ], - "execution_count": null - }, - { - "metadata": { - "tags": [] - }, - "source": [ - "### Interpretation of Discriminative Connectivity Patterns\n", - "\n", - "This plot shows the **most discriminative ROI connections** for classifying ASD vs Control subjects.\n", - "- **Red edges** indicate connections **stronger in ASD**.\n", - "- **Blue edges** indicate connections **stronger in Control**.\n", - "- Color intensity reflects the **magnitude of contribution** to the model’s decision.\n", - "\n", - "---\n", - "\n", - "**Key Patterns**:\n", - "\n", - "- **Default Mode Network (DMN)**:\n", - " - *DefaultMode.MPFC*, *DefaultMode.PCC*, *DefaultMode.LP (L/R)*\n", - " - Core hubs of the DMN, associated with **self-referential processing**, **social cognition**, and often disrupted in ASD.\n", - "\n", - "- **Fronto-Parietal Network**:\n", - " - *FrontoParietal.PPC (L)*\n", - " - Involved in **executive function** and **cognitive flexibility**, domains typically impaired in ASD.\n", - "\n", - "- **Dorsal Attention Network**:\n", - " - *DorsalAttention.IPS (L)*\n", - " - Associated with **goal-directed attention**, potentially altered in ASD subjects.\n", - "\n", - "- **Salience Network**:\n", - " - *Salience.SMG (R)*\n", - " - Plays a role in **interoception** and **social-emotional processing**, relevant for ASD symptoms.\n", - "\n", - "- **Language Network**:\n", - " - *Language.pSTG (R)*\n", - " - Critical for **language comprehension** and **social communication**, often affected in ASD.\n", - "\n", - "- **Sensorimotor and Cerebellar Regions**:\n", - " - *SensoriMotor.Lateral (L)*, *Cerebellar.Posterior*\n", - " - Linked to **motor coordination** and **sensorimotor integration**, commonly atypical in ASD.\n", - "\n", - "The interpretability analysis of the trained model highlights that **functional connectivity alterations across DMN, attention, salience, language, and sensorimotor systems** are key discriminative factors for distinguishing **ASD** from **Control** subjects." - ], - "cell_type": "markdown" - } - ] + ] } diff --git a/tutorials/brain-disorder-diagnosis/preprocess.py b/tutorials/brain-disorder-diagnosis/preprocess.py index 57779f5..9ea88a1 100644 --- a/tutorials/brain-disorder-diagnosis/preprocess.py +++ b/tutorials/brain-disorder-diagnosis/preprocess.py @@ -123,18 +123,19 @@ def preprocess_phenotypic_data(data, standardize=False): @validate_params( - {"data": ["array-like"], "measures": [list]}, prefer_skip_nested_validation=False + {"data": ["array-like"], "measures": [list, tuple]}, + prefer_skip_nested_validation=False, ) def extract_functional_connectivity(data, measures=["pearson"]): """Extract functional connectivity features from time series data. Parameters ---------- - data : list[array-like] of shape (n_subjects,) + data : list[array-like] or tuple[array-like] of shape (n_subjects,) An array of numpy arrays, where each array is a time series of shape (t, n_rois). The time series data for each subject. - measures : list[str], optional (default=["pearson"]) + measures : list[str] or tuple[str], optional (default=["pearson"]) A list of connectivity measures to use for feature extraction. Supported measures are "pearson", "partial", "tangent", "covariance", and "precision". Multiple measures can be specified as a list to compose a higher-order measure.