On the benefits of self-taught learning for brain decoding

TL;DR

This paper demonstrates that self-taught learning using large neuroimaging databases enhances brain decoding accuracy by pre-training models that learn more generalizable features, especially beneficial with limited labeled data.

Contribution

It introduces a self-taught learning framework leveraging neuroimaging databases to improve brain decoding performance and model generalization.

Findings

Self-taught learning improves classifier performance.

Benefits depend on data quantity and task complexity.

Pre-trained models are more robust to individual differences.

Abstract

Context. We study the benefits of using a large public neuroimaging database composed of fMRI statistic maps, in a self-taught learning framework, for improving brain decoding on new tasks. First, we leverage the NeuroVault database to train, on a selection of relevant statistic maps, a convolutional autoencoder to reconstruct these maps. Then, we use this trained encoder to initialize a supervised convolutional neural network to classify tasks or cognitive processes of unseen statistic maps from large collections of the NeuroVault database. Results. We show that such a self-taught learning process always improves the performance of the classifiers but the magnitude of the benefits strongly depends on the number of samples available both for pre-training and finetuning the models and on the complexity of the targeted downstream task. Conclusion. The pre-trained model improves the classification performance and displays more generalizable features, less sensitive to individual differences.