Detection of PTSD with Voice Analysis Software: A Systematic Literature Review

This work was sponsored by Boston VA Healthcare System Post-traumatic stress disorder (PTSD) is a prevalent and chronic mental health condition negatively impacting a person's psychological, occupational, physical, interpersonal and social functioning. Currently, PTSD diagnosis is determined by clinical interview or self-report assessment, which remain highly subjective and prone to biases. This has led to efforts to develop an objective and cost-efficient method to screen and diagnose a range of mental health conditions, including PTSD. Through voice analysis software, researchers have begun to explore an individual’s digital phenotype to distinguish specific voice features of a psychiatric disorder. The aim of this project was to review and synthesize the literature surrounding voice analysis software and its implications for improving PTSD detection. Review of the literature was conducted by querying PubMed and PsycINFO with Boolean search terms including PTSD or post-traumatic stress disorder, speech or voice analysis and detection, monitoring or screening. Of the 217 articles returned, 5 were selected for this review based on criteria and relevance to topic. Review of preliminary studies demonstrated significant evidence that voice analysis software could provide a reliable, valid, accessible, and objective method for screening and detecting PTSD. In summary, objective AI voice analysis software has demonstrated potential to provide a streamlined, non-intrusive and effective means for the screening and detection of mental health disorders including PTSD.