Project Spotlight: Multimodal Biomarkers (update)

Professor Louis-Philippe Morency, Finmeccanica Associate Professor of Computer Science in the Language Technology Institute at Carnegie Mellon University, and Dr. Eva Szigethy, Director of Behavioral Health within the Chief Medical and Scientific Office at UPMC and Professor of Psychiatry, Pediatrics and Medicine at the University of Pittsburgh, worked together to develop sensing technologies to automatically measure subtle changes in individuals’ behavior that are related to affective, cognitive, and psychosocial functioning. Their goal of the project was to develop and refine computational tools that automatically measure depression-related behavioral biomarkers and to evaluate the clinical utility of these measurements. We had a chance to connect with the team for an update and found out that most recently, they have extended their research to focus on language as well.

What was most exciting or unexpected that you learned during the project?

The focus of the research effort has been on analyzing behavioral markers related to depression. The team focused on behavioral markers related to language – how a patient expresses themselves. More specifically, they studied change in language when speaking with their doctors. In collaboration with doctors at UPMC, the team recorded 266 therapy sessions between 39 unique patients and 11 unique therapists (the original study was NSF funded). They recently analyzed these doctor-patient interactions to identify changes in language use. One of the main results was the link between patient’s language and their perception of the working alliance (the collaborative aspect of the doctor-patient relationship). A research paper summarizing these results was submitted to the International Conference on Multimodal Interaction (ICMI 2022) conference.

Did you encounter any roadblocks that you did not anticipate?

The team encountered a few unanticipated roadblocks. 1) Covid-19 caused the study to pause and the team had to reevaluate how to conduct the study remotely for participants; 2) there was turnover of behavioral health clinicians within the FGID clinic, who were the clinicians who recruited most of the patients to the study; 3) the RA could no longer go into clinic for recruitment purposes due to Covid-19, and clinic work flow changed during this time so that providers were primarily focused on the care of their patients in the short amount of time that they had to meet with them rather than referring them to this study.

What are your project’s next steps?

At UPMC, the team took advantage of another opportunity that Covid-19 presented in learning about how digital behavioral tools can enhance clinical care. They are now studying Long-COVID and how to improve cognitive fog and emotional distress, which are common and distressing symptoms for these patients. These PHDA seed funds were used to get pilot data in this population, which in turn helped Dr. Szigethy and the Long-COVID pulmonary clinic team receive a Beckwith Institute Clinical Transformation Program Research Grant. This randomized controlled study will evaluate the effectiveness of two strategies for treating patients with moderate or severe cognitive fog resulting from post-COVID symptoms, or long-COVID- stimulant medication or digital cognitive brain therapy with coaching using a mobile app to set the stage for a larger adaptive design trial.