Comment on Using voice analysis, AI diagnoses type 2 diabetes in 89% of women, 86% of men
Cheesus@lemmy.world 7 months ago
The auditory features that the AI focussed on included slight changes in pitch and intensity, which human ears cannot distinguish. This was then paired with basic health data gathered by the researchers, such as age, sex, height and weight.
Not quite just voice but still impressive
kozy138@lemm.ee 7 months ago
I bet age, weight, and height were probably bigger factors than the voice. Also larger people probably have a lower voice.
Windex007@lemmy.world 7 months ago
Yeah there is really solid, large scale studies that associate diabetes incidence against sex, age, and BMI. Really tight CI on the data.
dmention7@lemm.ee 7 months ago
When this was posted a few weeks ago, it was also pointed out that your socioeconomic status and geographical location can probably also be predicted pretty well by voice, which also correlate with T2D risk factors.
It is a neat finding and all, but unless there is some underlying physiological tie between T2D and your voice that’s not explainable otherwise, it’s kind of the diagnostic equivalent of “psychic” cold reading.