Is Detecting Diseases Based on 45 s of Voice Accurate? (Henry O'Connell)

Ambient speech technology is now common for documentation; but what if the biggest opportunity isn’t transcription? In this episode of Faces of Digital Health, host Tjasa Zajc speaks with Henry O’Connell (CEO of Canary Speech) about voice biomarkers: how conversational speech can support detection and monitoring of neurological and behavioral health conditions—without relying on “what you say,” but on how the nervous system produces speech. We cover: - Why decades of word-based voice biomarker research struggled to reach clinical practice Canary’s approach: analyzing the “primary data layer” behind speech production - How ~45 seconds of conversational speech can generate millions of data elements for analysis - Reported accuracy differences between progressive neurological diseases and behavioral health - Why validation must be repeated across languages and populations - Real-world use in primary care screening, clinical trials, and workflow-integrated ambient systems - Ethics and guardrails for wellness and workplace use - Emerging operational use cases, including aggression risk signals in hospital rooms Notes: 00:12 Intro: voice biomarkers beyond ambient documentation 01:03 Potential vs reality: what voice can (and can’t yet) prove clinically 01:54 Why earlier voice-biomarker work focused on words—and why it stalled 06:34 The “intuition” problem: we hear mood without words 07:33 Canary’s clinical-first approach and global clinical partnerships 08:39 Do you need long-term data? (Agatha Christie example) 09:41 Method: ~45 seconds of conversational speech, ambient capture 10:36 Scale: 2,590 features every 10ms (~13M data elements) 11:02 Where the features come from (vocal cords, respiration, resonance) 14:36 How models are built: IRB, clinician ground truth, ML correlation 17:30 Accuracy + adoption: why it’s not standard practice yet 18:22 Reported performance: 98%+ neuro, ~80s behavioral health 22:44 Culture/language bias concern: why validation per language matters 24:22 Guardrails: validate every new language and population testing 31:07 Why read-speech scripts misled the field; conversational-only stance 34:16 What changed: AI, compute, real-time streaming, workflow fit 37:10 How it’s used: screening vs suspected disease; incidental findings 39:24 Primary care example: postpartum depression flagged despite “I’m fine” 47:56 Wellness/employee use: de-identified dashboards and ethics 53:42 Technical requirements: device capture, signal-to-noise checks 57:10 In-room monitoring: aggression risk signals for staff safety 1:02:22 Clinical trials: pre-screening + measuring therapeutic impact 1:07:14 Global rollout: regions, languages, and partnerships 1:09:42 Consumer access: wellness product planned in 2026 1:12:32 Wrap-up: why this matters as cognitive health needs grow