Whole MILC: generalizing learned dynamics across tasks, datasets, and populations

TL;DR

This paper introduces whole MILC, a self-supervised learning method that captures whole sequence mutual information in fMRI data, enabling effective cross-task, dataset, and population generalization for disorder diagnosis.

Contribution

The paper presents a novel self-supervised training schema, whole MILC, which pre-trains on unlabeled data to improve disorder classification and interpretability across diverse datasets.

Findings

Outperforms existing self-supervised methods in disorder classification.

Provides competitive results compared to classical machine learning algorithms.

Enables attribution of diagnoses to specific spatio-temporal brain regions.

Abstract

Behavioral changes are the earliest signs of a mental disorder, but arguably, the dynamics of brain function gets affected even earlier. Subsequently, spatio-temporal structure of disorder-specific dynamics is crucial for early diagnosis and understanding the disorder mechanism. A common way of learning discriminatory features relies on training a classifier and evaluating feature importance. Classical classifiers, based on handcrafted features are quite powerful, but suffer the curse of dimensionality when applied to large input dimensions of spatio-temporal data. Deep learning algorithms could handle the problem and a model introspection could highlight discriminatory spatio-temporal regions but need way more samples to train. In this paper we present a novel self supervised training schema which reinforces whole sequence mutual information local to context (whole MILC). We pre-train the whole MILC model on unlabeled and unrelated healthy control data. We test our model on three different disorders (i) Schizophrenia (ii) Autism and (iii) Alzheimers and four different studies. Our algorithm outperforms existing self-supervised pre-training methods and provides competitive classification results to classical machine learning algorithms. Importantly, whole MILC enables attribution of subject diagnosis to specific spatio-temporal regions in the fMRI signal.