Charles Marmar, MD, chair and director of the Department of Psychiatry – PTSD Research Program at NYU Langone Health, shares details about the potential to leverage artificial intelligence and machine learning, particularly the random forest algorithm, in the analysis of hi-fidelity audio recordings of patients to comb through “40,000 unique biophysical features” in order to uncover any of the 18 features the research team has associated with post-traumatic stress disorder (PTSD).

Dr Marmar’s colleague, Eugene Laska, PhD, expands upon the mechanisms of the random forest algorithm and how it may assist psychiatrists in determining the probability of a certain diagnosis and, subsequently, confirming or disproving the suspected diagnosis.