Towards Artificial Intelligence in Mental Health by Improving Schizophrenia Prediction with Multiple Brain Parcellation Ensemble-Learning
Kalmady SV, Greiner R, Agrawal R, et al. Towards artificial intelligence in mental health by improving schizophrenia prediction with multiple brain parcellation ensemble-learning. npj Schizophrenia. 2019;5(1):2. https://doi.org/10.1038/s41537-018-0070-8.
In the literature, there are substantial machine learning attempts to classify schizophrenia based on alterations in resting-state (RS) brain patterns using functional magnetic resonance imaging (fMRI). Most earlier studies modelled patients undergoing treatment, entailing confounding with drug effects on brain activity, and making them less applicable to real-world diagnosis at the point of first medical contact. Further, most studies with classification accuracies >80% are based on small sample datasets, which may be insufficient to capture the heterogeneity of schizophrenia, limiting generalization to unseen cases. In this study, we used RS fMRI data collected from a cohort of antipsychotic drug treatment-naive patients meeting DSM IV criteria for schizophrenia (N = 81) as well as age- and sex-matched healthy controls (N = 93). We present an ensemble model — EMPaSchiz (read as ‘Emphasis’; standing for ‘Ensemble algorithm with Multiple Parcellations for Schizophrenia prediction’) that stacks predictions from several ‘single-source’ models, each based on features of regional activity and functional connectivity, over a range of different a priori parcellation schemes. EMPaSchiz yielded a classification accuracy of 87% (vs. chance accuracy of 53%), which out-performs earlier machine learning models built for diagnosing schizophrenia using RS fMRI measures modelled on large samples (N > 100). To our knowledge, EMPaSchiz is first to be reported that has been trained and validated exclusively on data from drug-naive patients diagnosed with schizophrenia. The method relies on a single modality of MRI acquisition and can be readily scaled-up without needing to rebuild parcellation maps from incoming training images.
Consistently altered expression of gene sets in postmortem brains of individuals with major psychiatric disorders
Darby M,M., Yolken R,H., Sabunciyan S. Consistently altered expression of gene sets in postmortem brains of individuals with major psychiatric disorders. Transl Psychiatry. 2016;6:e890. http://dx.doi.org/10.1038/tp.2016.173.
The measurement of gene expression in postmortem brain is an important tool for understanding the pathogenesis of serious psychiatric disorders. We hypothesized that major molecular deficits associated with psychiatric disease would affect the entire brain, and such deficits may be shared across disorders. We performed RNA sequencing and quantified gene expression in the hippocampus of 100 brains in the Stanley Array Collection followed by replication in the orbitofrontal cortex of 57 brains in the Stanley Neuropathology Consortium. We then identified genes and canonical pathway gene sets with significantly altered expression in schizophrenia and bipolar disorder in the hippocampus and in schizophrenia, bipolar disorder and major depression in the orbitofrontal cortex. Although expression of individual genes varied, gene sets were significantly enriched in both of the brain regions, and many of these were consistent across diagnostic groups. Further examination of core gene sets with consistently increased or decreased expression in both of the brain regions and across target disorders revealed that ribosomal genes are overexpressed while genes involved in neuronal processes, GABAergic signaling, endocytosis and antigen processing have predominantly decreased expression in affected individuals compared to controls without a psychiatric disorder. Our results highlight pathways of central importance to psychiatric health and emphasize messenger RNA processing and protein synthesis as potential therapeutic targets for all three of the disorders.