Identifying PTSD by testing voices of artificial intelligence

A new study found that a specially designed computer programme can help veterans diagnose PTSD by testing their voices. The study was published in Depression and Anxiety newspaper. In the study, an AI (artificial intelligence) tool can differentiate between voices of persons with or without PTSD, with almost 89 percent accuracy. “Our results indicate that speech-based characteristics can be used to identify the disease, and that with further modification and validation, they can be used in the clinic for the coming period.” Charles R. Marmar—Senior Study Author and Chairman of the Department of Psychiatry of the NYU School of Medicine.

More than 70% of the world’s adults experience traumatic events at certain points in their lives, with at least 12% of people suffering from PTSD in some fighting countries. Those with the medical condition suffer persistent and severe distress when they recall an event. A self-report evaluation or a clinical interview is the most common way of deciding PTSD identifiers, both of which are naturally biassed. This led to efforts to achieve objective, measurable physical PTSD progression indicators, similar to laboratory values under medical circumstances, but slow improvements. The study utilised a method for statistical and machine learning — so-called random forests — that enables the research group to “learn” how people could be categorised based on examples.

A recent study found a connection between PTSD use of healing and compensation tests on the same note. According to a recent study by the department of psychiatry researchers of YU, veterans who wanted job-related PTSD compensation were more likely to take PTSD-correlated treatment sessions before their compensation examinations than later, but only when experts were convinced that there would be a combination of treatment compensation. PLOS One published this study.

Related Posts

Leave a Reply

Your email address will not be published. Required fields are marked *