Tulay Adali, fourth from left, with the members of her lab. Photo courtesy of Adali.

UMBC and Georgia State receive $3M NIMH grant to improve data-driven diagnosis of mood disorders

UMBC and Georgia State University have received a $3 million five-year grant from the National Institute of Mental Health (NIMH) for research supporting the diagnosis of mood disorders. Tulay Adali, professor of computer science and electrical engineering (CSEE) and distinguished university professor, will lead UMBC’s portion of the research, which will receive about $870,000 in support.

Mental illnesses and mood disorders are complicated and can be challenging to identify, says Adali. Diagnoses are often made based on symptoms that a person experiences, rather than using quantifiable measures, and descriptions of symptoms can be quite variable and subjectively observed and evaluated. 

The research team hopes to improve doctors’ ability to diagnose mood disorders through more quantitative, consistent measures. They will develop dynamic approaches to understanding how the continuously changing state of the brain is affected by mental illness. And their recommendations will include data from a range of sources, to more accurately reflect the complexity of mental illness.

Adali will work with her former graduate student Vince Calhoun, Ph.D. ‘02, electrical engineering. Calhoun is currently the director of the Center for Translational Research in Neuroimaging and Data Science (TReENDS) at Georgia State University. Adali and Calhoun have worked together on multiple research grants in the past. 

In this project, the UMBC group led by Adali will focus on diagnostic methods, particularly the use of medical imaging data, including functional magnetic resonance imaging (fMRI). Adali and her team will develop multivariate data-driven models to help capture changes over time and space. They will apply these models to large datasets to evaluate their performance as diagnostic tools. The researchers will assess the reproducibility and replicability of the methods that are developed.

“I am especially excited about our proposal to identify homogeneous subgroups of subjects in a completely data-driven manner from neuroimaging data,” says Adali. “We hope this will enable us to better define subtypes of mental disorders and will help inform effective and personalized forms of therapy.” 

This story was adapted from a UMBC News article written by Megan Hanks.