A+Voice is a research-stage programme collecting voice samples to study whether acoustic features carry usable health signal. Because this area attracts overclaiming, it is worth being precise about what the work is.
What is being studied
Speech carries measurable acoustic properties — timing, pitch variation, articulation stability, spectral characteristics. Research across several groups has explored whether changes in these properties correlate with certain neurological and respiratory conditions.
A+Voice contributes to that research question, collecting samples under consent with contributors retaining control over their data.
What this is not
A+Voice does not provide a validated diagnosis. Nothing produced by this programme should be used to make a clinical decision, and no output is a substitute for assessment by a qualified clinician.
Correlation observed in a research cohort is not a diagnostic test. The distance between "this acoustic feature differs across groups in our sample" and "this measurement can tell an individual something actionable about their health" is substantial, and it is measured in clinical validation studies that have not been completed.
Why the framing matters
Voice is attractive to work on precisely because collection is cheap and non-invasive. That same property makes it easy to ship something that looks like a health product long before the evidence supports one.
Holding research-stage framing is a deliberate constraint, not modesty.
Data control
Contributors decide whether their samples are included, and can withdraw. Samples are handled as identifiable data — voice is not anonymous.