Leveraging Large-scale Data to Improve Psychiatric Care

Jordan W. Smoller, MD, ScD, is the director of the Psychiatric and Neurodevelopmental Genetics Unit and is the associate chief for research in the Department of Psychiatry at Mass General. In this video, he discusses his research efforts to leverage large-scale data sets that are now available from a variety of sources, including neuroscience, genomics and electronic health records, to improve the ability to diagnose, detect and treat psychiatric disorders and serious complications of those disorders.

Subscribe to the latest updates from Neuroscience Advances in Motion

Thank you for subscribing!

Error: Please enter a valid email address.

Email Address Submit

Transcript

Can we leverage large-scale data that are now available in the realm of neuroscience, genomics, electronic health records to improve our ability to diagnose, detect and treat psychiatric disorders and serious complications of those disorders?

We've been doing quite a bit of work using large-scale brain imaging datasets combined with genomics to understand how genes affect the structure and function and connectivity of brain circuitry. How is that related to behavior and psychiatric disorders?

So, we have, for example, looked at ways in which genetic loading for psychiatric disorders may influence the connectivity at a structural or functional level in different circuits and regions of the brain. So to give you an example, we have found that genetic influences on the structure and thickness of regions of the cortex in specific areas, particularly around the superior temporal gyrus, seem to be related to genetic influences on schizophrenia.

We've also found that genetic influences on depression proneness seem to play out to some degree in the structural connectivity of regions of the prefrontal cortex that modulate amygdala function, and those are giving us clues about some of the specific circuits that may be involved in those conditions.

Another area which is becoming really a focus of research and has really only been possible in the last several years is to leverage some of these very large-scale data sets that are available through, for example, electronic health records, biobanks and genomic resources to see whether we can make progress in predicting some of the outcomes and onset of disease for which we really have no robust predictors.

For example, we've used electronic health records and artificial intelligence, or machine learning approaches, to see whether we could develop a predictive algorithm for suicide risk. We're now in the process of trying to validate those algorithms in health care systems around the country to see whether we might be able to develop a clinical decision support tool that could assist clinicians at the point of care and hopefully bend the curve for an outcome, which is hard to predict but catastrophic.

We're really trying to build tools that can make a difference in clinical care and address questions for which clinicians have not had really an ability to identify the course, prognosis or risk in many areas of psychiatry, whether it's the risk of post-traumatic stress disorder after trauma, the risk of psychotic illness among young people or whether somebody might respond to a given course of treatment. There are many possibilities, and we're very excited about this line of work.

Learn more about the Psychiatric and Neurodevelopmental Genetics Unit

Learn more about research in the Department of Psychiatry