Approved Research

Redefining Trans-Neuropsychiatric Disorder Patterns through Big Data and Machine Learning

Lay summary

We aim to develop and validate novel brain-based measurements that will advance research and guide personalized medical care in mental and neurological illnesses. Our project uses Big Data derived brain deficit patterns to quantify the similarity between an individual's brain and the brain patterns observed in common brain disorders such as depression, schizophrenia, Alzheimer's dementia, Parkinson's disorder, and others. In the past we demonstrated that Brain Deficit Patters A)  develop 1-3 decades prior to the onset of symptoms. B) show excellent replicability in independent cohorts; C) show strong overlap across related illnesses e.g. mild cognitive impairment and Alzheimer's dementia and D) may guide medical decisions including diagnosing the stage, predicting treatment resistance, and estimating risks for progression. The Big Data brain deficit patterns serve as the basis for comparison that computes "similarity to expected deficit patterns", similar to Polygenic Risk Scores (PRS) which are used to compute genetic vulnerability for the illness. When compared to PRS, the brain-based measurements capture both genetic and environmental risk factors and show ~2-10 times higher diagnosis-related variance than PRS; are more informative of cognitive variance, symptom severity, and treatment resistance; and are not sensitive to ethnic differences, allowing the use of multi-ethnic cohorts. Our hypothesis is that the formation of disorders' specific brain patterns precedes the development of clinical symptoms and therefore can be used as early diagnostic tools for illnesses such as dementia, where early diagnosis is key. We will test these hypotheses in UKBB by performing patient-control analyses in the corresponding illness groups; studying the degree of separation vs. commonality across disorders, and evaluating pattern differences at specific stages of the illnesses in cognitive scores and symptom severity.  We will use multivariate analyses to link the brain-based measurements to variance in cognitive domains ascertained by UKBB. This project is expected to last about 5 years and will benefit the community at large by populating the UKBB database with novel brain measurements extracted and homogenized using standard workflows that build on Open Science approaches.

Scope extension:

We will use data from two large scale  initiatives (Connectomes Related to Human Disease (CRHD) and Enhancing Neuroimaging Genetics through Meta-Analysis (ENIGMA)  to understand disorder-related patterns in the human brain. We will UKBB sample as an independent cohort to evaluate sensitivity and specificity of novel brain vulnerability metrics - based on the idea of polygenic risk scores - that we expect to better predict diagnosis and cognitive performance than standard neuroimaging measures. We define a metric of "vulnerability" by quantifying the similarity between controls' brain patterns and deficit patterns in neuropsychiatric disorders. The Regional Vulnerability Index (RVI) uses Big Data meta-analyses to quantify the similarity between an individual and meta-analytical deficit effect size patterns based on large and diverse international samples. The Machine Learning-Vulnerability Index (MVI) is trained using Big Data mega-analytic samples to quantify the similarity for individual brain patterns to those learned from patients and controls. We will compute novel, brain-based metrics to provide a measurement for each of the UKBB individuals with neuroimaging data with similarity index to: schizophrenia-spectrum and psychosis disorder, major depression,  bipolar disorder, epilepsy, mild cognitive impairment, and Alzheimer's disease.

We would like to expand the scope to include the study of the effects of chronic metabolic illnesses, such as hypertension, hyperlipidemia, and hyperglycemia on the brain in the context of neuropsychiatric research. We hypothesize that common genetic factors may underly the brain-body relationship in people with mental illnesses and the higher rates of metabolic illness further contribute to the brain deficits.