Machine learning for subtype identification and diagnosis of depression from multimodal data

[Speaker] Yoshimoto, Junichiro:1,2
1:Nara Institute of Science and Technology (Japan), 2:Okinawa Institute of Science and Technology (Japan)

Diagnosis of depression is currently based on patients' reports on depressive mood and loss of motivation. In the search for an objective and more efficient method to diagnose this disorder, we analyzed multimodal data, including MRI, blood markers, genetic polymorphism and behavioral tests, obtained from depression patients and healthy controls, using supervised and unsupervised machine learning algorithms. For the supervised learning, we applied L1-regularized logistic regression to diagnose patients and predict the chance of recovery after treatment with SSRI. Verified through cross validation, the depression diagnosis and the treatment effect prediction could be achieved with a high accuracy. For the unsupervised learning, we developed a novel Bayesian co-clustering algorithm. Using the algorithm, we found three sub-clusters in depression patients. More interestingly, resting-state functional connectivities associated with the clusters were in the default-mode network.
Advanced Search