Potential Vocal Tracking App Could Proactively Identify Depression and Schizophrenia



Original Source Here

Harking back to the original topic, MDD and schizophrenia are reportedly some of the more common precursors to suicide. If we follow this line of thinking, then proactive measures are needed to identify MDD and/or schizophrenia quickly.

The Acoustical Society of America (ASA) is an international scientific society that specializes in promoting, disseminating, and generating knowledge of acoustics and its practical applications.

Acoustics, in this case, pertains to the field of physics that deals with the production, reception, control, and transmission of mechanical waves, including vibration, sound, infrasound, and ultrasound.

Each year, the ASA has their usual meeting. This year, Dr. Carol Espy-Wilson, from the University of Maryland, is planning to discuss how a person’s mental health status reflects their speech lectures on Tuesday, June 8.

Dr. Carol Espy-Wilson and her former graduate student, Dr. Ganesh Sivaraman have done a lot of research surrounding speech inversion mapping.

If we follow the trail here, this means that speech inversion systems could map out acoustic signals. If we create a robust form of a speech inversion map, then it could easily detect many things such as:

  • Changes in general language learning
  • Automatic speech recognition
  • Detection of depression and/or schizophrenia in speech
  • Speech accent improvements

If this speech inversion map can track all these different variables down in high amounts across many different voices, then perhaps it would possible to determine if someone has depression or schizophrenia — just by their voice alone.

For example, someone with depression might talk slower, have a harder time thinking quickly on their feet, and have more pauses in their speech. Someone with schizophrenia might talk quickly or with little enthusiasm, with some loose associations between topics, with very unique phrases, and repetitive statements.

Thus, machine learning could be used to generate enough data surrounding vocal tone where proactive mental health classification is possible. If enough data is generated, it could easily be housed in a smartphone application that could help clients and physicians.

For now, though, we just need our eyes, mouths, and ears open. Perhaps in due time, this app will be fully fleshed out, and serve as another tool to combat serious mental health problems before their fully arise.

AI/ML

Trending AI/ML Article Identified & Digested via Granola by Ramsey Elbasheer; a Machine-Driven RSS Bot

%d bloggers like this: