You know me, I am always looking for aging developments or diagnostic tools to tell us how we’re doing with the process of growing older. To that end, I found a very interesting article that hit my fancy and interest. It is an every-which way piece that has a lot of tidbits.
Data from a CDC survey claims that 22% of all U.S. adults got into line for mental health treatment; up from 19% two years ago and, of course, it corresponds with the COVID19 pandemic.
What this is leading to, is that in the health-research department, there is a voice biomarker that is so AI-smart, it can recognize depression. We can generally tell when someone moves from cheerful to depressed. In my own assessment, in the late 90’s I interviewed some quadriplegics for a chapter in my book: Bringing Up Ziggy: What Raising a Helping Hands Monkey Taught Me About Love, Commitment, and Sacrifice. I asked them to tell me about their accident and how they got hurt. They were, of course, morose, and sad at the remembering. Then I asked them what their life was like with their monkey-helper. A complete change in tone, dialect, and now anxious to relate a story—they were ebullient in tone!
But now, apparently hospitals and insurance companies are taking it further by installing voice biomarker software that can identify if the caller is depressed. It is supposed to be with a patient’s permission, and I wonder how that conversation will go? “Hellooo, caller, we are listening to your voice to see if you need therapy.” One of the companies leading this charge is Kintsugi, where their lead-in is: “You see mental health differently when you listen between the lines.” Not only will vocal biomarkers be used for diagnosis but also for risk prediction and remote monitoring of clinical outcomes and symptoms.
In addition voice biomarkers are able to suss out neurodegenerative disorders such as Parkinson’s disease, Alzheimer’s, and mild cognitive impairment, in addition to multiple sclerosis, rheumatoid arthritis, and other cardiovascular diseases which can recognize jitters, voice weakness, articulatory impairment and such significance as psychoacoustic complexity to help pinpoint vocal impairment. A newsy brief written by Axios says that the vocal biomarkers analyze how the subject talks, taking into account the intonation, pitch, prosody—a metric rhythm—and it is sampled by an AI that compares it against a database of ‘anonymized voices’.
And it is not proof however, (just as evidence challenges in a courtroom,) but employers want it and companies that sell the technology are raising funds and soliciting clients to demonstrate its growing clinical usefulness. We will have to see how the privacy fallout concerns go down as well.
Voice for Health: The Use of Vocal Biomarkers from Research to Clinical Practice