An Emotion-Driven Vocal Biomarker-Based PTSD Screening Tool

Goal: This paper introduces an automated post-traumatic stress disorder (PTSD) screening tool that could potentially be used as a self-assessment or inserted into routine medical visits for PTSD diagnosis and treatment. xmlns:xlink="http://www.w3.org/1999/xlink">Methods: With emotion estimation algorithm providing arousal (excited to calm) valence (pleasure displeasure) levels through discourse, we select regions of the acoustic signal are most salient detection. Our was tested on subset data from DVBIC-TBICoE TBI Study, which contains Check List Civilian (PCL-C) assessment scores. xmlns:xlink="http://www.w3.org/1999/xlink">Results: Speech low-arousal positive-valence provide best discrimination PTSD. model achieved AUC (area under curve) equal 0.80 in detecting PCL-C ratings, outperforming models with no filtering (AUC = 0.68). xmlns:xlink="http://www.w3.org/1999/xlink">Conclusions: result suggests drives selection temporal audio recording