Jointly Discriminative and Generative Recurrent Neural Networks for Learning from fMRI
This work addresses data scarcity and interpretability issues in fMRI analysis for autism diagnosis, representing an incremental advance in multitask learning for medical imaging.
The authors tackled the challenges of limited fMRI data and unclear network interpretability by developing a novel RNN-based model that jointly learns to classify autism vs. healthy controls and generate fMRI time-series data, improving classification learning and producing meaningful functional communities.
Recurrent neural networks (RNNs) were designed for dealing with time-series data and have recently been used for creating predictive models from functional magnetic resonance imaging (fMRI) data. However, gathering large fMRI datasets for learning is a difficult task. Furthermore, network interpretability is unclear. To address these issues, we utilize multitask learning and design a novel RNN-based model that learns to discriminate between classes while simultaneously learning to generate the fMRI time-series data. Employing the long short-term memory (LSTM) structure, we develop a discriminative model based on the hidden state and a generative model based on the cell state. The addition of the generative model constrains the network to learn functional communities represented by the LSTM nodes that are both consistent with the data generation as well as useful for the classification task. We apply our approach to the classification of subjects with autism vs. healthy controls using several datasets from the Autism Brain Imaging Data Exchange. Experiments show that our jointly discriminative and generative model improves classification learning while also producing robust and meaningful functional communities for better model understanding.