A new study showing that voice analysis software can accurately diagnose posttraumatic stress disorder (PTSD) has created world-wide interest.
Diagnosing PTSD is a challenge. Diagnosis is usually based on clinical interviews or self‐report measures, however, the very nature of PTSD symptoms, along with stigma, form a significant barrier to accurate diagnosis, with patients liable to either under-report or exaggerate their symptoms. What is needed is an objective measure that can accurately diagnose PTSD.
A study of veterans conducted by the New York University School of Medicine has shown that voice analysis software was able to detect which individuals had PTSD and which ones did not with a high degree of accuracy.
Changes in speech quality, such as altered tone or agitated speech, are often used by clinicians as part of their assessment of individuals with mental health disorders.
The study team used an artificial intelligence software program that ‘learns’ how to classify individuals based on examples of speech. Using recorded patient interviews, the software linked patterns of specific voice features, including less clear speech and a lifeless, metallic tone, with PTSD in order to make a diagnosis.
“There is a growing trend to harness the power of machine learning to try and gain a more data-driven understanding of mental health disorders”.
Dr Tracey Varker, Senior Research Fellow at Phoenix Australia
Tracey adds that, “Phoenix Australia is currently involved in research investigating machine learning and biometrics in relation to PTSD diagnosis. It is an exciting area, and how far this new technology will take us is yet to be seen, but it has the potential to assist clinicians in making more accurate diagnoses and targeting treatment accordingly”.