The study of translational biomarkers in brain disorders is a very challenging and fruitful approach, which will empower a better understanding of healthy and diseased brains. This project will promote the translation of advanced engineering solutions and mathematic tools to novel neuroimaging applications in psychiatric disorders including major depression disorder(MDD), bipolar disorders(BD) and schizophrenia(SZ), allowing sophisticated and powerful analyses on highly complex datasets. To date, the unifying syndrome classification (ICD-9/10;DSM-IV/5) for these mental disorders obscures our knowledge of underlying pathophysiology and cannot guide optimal treatments. For example, there is no biomarker that is able to precisely predict response of MDD to some treatments. One reason lies in that most neuroimaging ?prediction? studies to date have used a single imaging measure or reported simple ?correlation? relationships, without considering multimodal cross- information, or lack of multi-site validation. Hence, developing novel data mining techniques such as deep learning, fusion with references, and sparse regression etc. can complement and exploit the richness of neuroimaging data, which can be promising avenues to identify objective biomarkers, which goes beyond a more descriptive use of brain imaging as traditionally used in studies of brain diseases. We will develop 3 novel data- driven methods: 1) A supervised fusion model that can provide insight on how cognitive impairment(in SZ) or epigenetic factors (miR-132 dysregulation in MDD) may affect covarying brain function and structure, which uses different clinical measures as reference to guide multimodal MRI fusion; 2) A cutting-edge prediction model that is able to identify imaging biomarkers for precise, individualized prediction of clinical outcomes, e.g., remission status of MDD patients after Electroconvulsive Treatment(ECT). 3) We will draw on advances and ideas from deep neural networks(DNN) combined with layer-wise relevance propagation (LRP), to classify multiple group of psychiatric disorders by using functional connectivity measures, and to trace back the most predictive functional networks from the black box of deep learning by LRP. All above proposed methods will be applied to the big data containing multimodal imaging and behavioral information(n=5000) pooled from existing studies, to investigate biomarkers that can help solve specific clinical difficulties. This pioneering study may provide an urgently-needed paradigm shift in the treatment and diagnosis of psychiatric disorders, thereby guiding personalized clinical care. Accomplishment of this project has great potential to discover neuroimaging biomarkers that have been missed by existing approaches, lead to earlier and more effective interventions, suggesting a significant translational impact.
Psychiatric imaging is struggling with identifying robust biomarkers. Existing approaches do not fully leverage the power of multimodal data, despite evidence that such information is highly informative. We will draw on advances and ideas from fields of deep neural networks, supervised learning and dynamic functional information, to capture rich information from multimodal imaging big data, thus identify replicable and precise biomarkers that are able to predict individual clinical measures and help for intervention. We will pool big data from ongoing projects in multiple cohorts, consisting a large imaging and behavioral dataset (n~5000) to apply clinical applications that will have profound translational medicine impact on schizophrenia, bipolar disorder and major depressive disorders.
