A Zero-shot 2D Sentiment Model Predicts Clinical Outcome in Psilocybin Therapy for Treatment Resistant Depression

Abstract: 

Background: Therapeutic administration of psychedelic drugs has shown the potential to improve mental health in many historical anecdotal accounts as well as in an extensive and growing scientific literature. A recent randomized double-blind phase-IIb study demonstrated the safety and efficacy of psilocybin in participants with treatment resistant depression (TDR). While promising, this study also showed that the treatment works for a portion of the TRD population, and thus early prediction of outcome is a key objective.

Methods: Transcripts were made from audio recordings of the psychological support sessions one day post psilocybin administration. We used a zero-shot machine learning classifier based on the BART large language model fine-tuned on the Multi-Natural Language Inference dataset to compute two-dimensional sentiment (valence and arousal) for the participant and therapist from the transcript. These scores, combined with the Emotional Breakthrough Index (EBI) and treatment arm were used to predict treatment outcome as measured by MADRS scores. Code and data are available at https://github.com/compasspathways/Sentiment2D

Results: Two multinomial logistic regression models were fit to predict responder status at week 3 and through week 12. Cross-validation of these models resulted in 85% and 88% accuracy and AUC values of 88% and 85%.

Conclusions: A machine learning algorithm using NLP and EBI accurately predicts long term patient response, allowing rapid prognostication of personalized response to psilocybin treatment and insight into therapeutic model optimization. Further research is required to understand if language data from earlier stages in the therapeutic process hold similar predictive power.

Bio: 

As the Vice President of Digital Health Research at COMPASS Pathways, Bob is leading the data science and machine learning efforts aimed at improving the safety, efficacy, and scalability of psilocybin therapy. He is an accomplished neuroscientist and engineer with deep expertise in measuring human brain and behavior, and building data-driven solutions to mental health care challenges. Prior to joining COMPASS Pathways, Bob was VP of Research at Mindstrong, leading the research and data science teams in the development of digital biomarkers for mental health. Prior to Mindstrong, Bob was the Research Director of the Stanford Center for Neurobiological Imaging. He has published over one hundred peer-reviewed articles in the fields of psychology, psychiatry, neuroscience, statistics, and magnetic resonance technology over his 30+ year scientific career. Bob completed his PhD in Experimental Psychology at the University of California at Santa Cruz, and postdoctoral fellowships at the University of British Columbia and Stanford University.

Open Data Science

 

 

 

Open Data Science
One Broadway
Cambridge, MA 02142
info@odsc.com

Privacy Settings
We use cookies to enhance your experience while using our website. If you are using our Services via a browser you can restrict, block or remove cookies through your web browser settings. We also use content and scripts from third parties that may use tracking technologies. You can selectively provide your consent below to allow such third party embeds. For complete information about the cookies we use, data we collect and how we process them, please check our Privacy Policy
Youtube
Consent to display content from - Youtube
Vimeo
Consent to display content from - Vimeo
Google Maps
Consent to display content from - Google