What we are. What we are not.
digitalhumanOS™ is a clinical screening platform for multimodal screening of speech and visual biomarkers. The platform's clinical claims are grounded in 19 peer-reviewed published studies. This page describes the regulatory framework we operate within, including what FDA establishment registration is and is not.
Key Facts
- Clinical Validation
- Peer-reviewed (19 studies)
- Conditions
- 46
- Cognitive Decline AUC
- 0.890
- Compliance
- HIPAA · SOC 2 Type II
- Deployment
- Samsung Knox
Peer-reviewed across 19 published studies. HIPAA compliant. Human-in-the-loop by design. Every result reviewed by a licensed clinician before it enters the clinical record.
This content is intended for informational purposes and does not constitute medical advice. Editorially reviewed by David Kaiser, CEO of Scienza Health, for accuracy in post-acute care operations.
Establishment registration is administrative, not an endorsement
FDA establishment registration under 21 CFR Part 807 is an annual administrative filing that medical device facilities make with the FDA. It is not a review of any specific device, and it is not the same as 510(k) clearance or PMA approval.
Per 21 CFR 807.39, “registration of a device establishment or assignment of a registration number does not in any way denote approval of the establishment or its products by FDA.” We disclose registration status here as a matter of transparency, not as a substitute for independent peer-reviewed clinical evidence — which is the appropriate standard for evaluating any AI screening tool.
Peer-Reviewed. Independently Validated.
Our screening models aren't validated on curated lab data. They are validated against real clinical outcomes from real patients in real care settings. Depression screening at AUC 0.816 (Journal of Affective Disorders). PTSD at AUC 0.800 (Journal of Traumatic Stress). Anxiety at AUC 0.775 (Psychological Medicine). Parkinson's at an AUC of 0.97 (Movement Disorders). Cognitive decline at AUC 0.890. These numbers come from peer-reviewed clinical research by independent researchers at Beth Israel Deaconess Medical Center, NIH, MIT, Harvard Medical School, and The Lancet — not marketing claims.
GIA® screens. Clinicians decide.
GIA®, powered by digitalhumanOS™, is a screening tool — not a diagnostic device. She flags potential conditions for clinician review. Every result is reviewed and approved by a licensed clinician before it enters the permanent clinical record. The system assists clinical judgment. It never replaces it. This is the same standard that applies to MoCA, MMSE, BIMS, and every other clinical screening instrument.
digitalhumanOS™ does not diagnose, treat, cure, or prevent any disease or condition. All screening results require review and clinical judgment by a qualified healthcare professional before any clinical action is taken.
For medical directors and compliance officers evaluating GIA® for clinical deployment. See how GIA® integrates with PointClickCare, Epic, Cerner, and MatrixCare.
Screening, Not Diagnosis
Every flagged condition requires clinician confirmation. The system identifies what needs attention — the clinician decides what to do about it.
Continuous Monitoring
Model performance is monitored continuously against real-world outcomes. Accuracy doesn’t degrade because we catch it before it does.
Transparent Methodology
Our clinical validation methodology, accuracy metrics, and limitations are published and peer-reviewed. We don’t ask for trust. We earn it.