Whitney Health
Clinical Evidence & Validation
Built on a decade of peer-reviewed research, NIH-funded clinical validation, and priority FDA regulatory engagement.
Whitney Health delivers automated, self-administered digital assessment that substantially outperforms standard cognitive screens in diagnostic accuracy across mild cognitive impairment subtypes and late-life depression. By providing objective differentiation between conditions that present similarly but require fundamentally different management, the platform gives health systems the clinical evidence base to optimize patient selection, support therapeutic safety, and reduce the risks of diagnostic error in an increasingly complex neuro-therapeutic environment.
Regulatory Status & Clinical Pathway
The program combines FDA priority engagement with a defined pathway toward commercially cleared differential diagnosis.
FDA Breakthrough Device Designation
Granted for automated diagnosis and differential diagnosis of MCI and assessment of neurocognitive, neurobehavioral, and psychiatric function in adults aged 65–84.
De Novo Submission in Preparation
The platform is in preparation for a De Novo submission for automated differential diagnosis of amnestic MCI, non-amnestic MCI, and late-life depression.
First-in-Class Specificity
Upon clearance, Whitney Health will be the first commercially cleared digital tool for automated differential diagnosis across these three conditions within one 60-minute self-administered session.
Patient Safety: De-Risking Anti-Amyloid Therapy Selection
Amnestic MCI, non-amnestic MCI, and late-life depression frequently present with overlapping clinical symptoms. They demand fundamentally different management strategies. The inability to distinguish between them is not a minor inconvenience — in the current treatment environment, it is a source of significant clinical, financial, and regulatory risk.
The clinical imperative
Patients with amnestic MCI may be appropriate candidates for disease-modifying anti-amyloid therapies. Patients with late-life depression are not.
The safety risk
Prescribing anti-amyloid agents to misdiagnosed patients exposes individuals to unnecessary clinical risk, including ARIA, while creating financial and regulatory liability.
Objective separation
Whitney Health separates diagnostic ambiguity through interpretable sub-indices derived from raw behavioral features, including response speed, variability, and speech-language measures.
General response speed separates aMCI and naMCI from healthy controls with an AUROC of 0.84.
Late-life depression is separated by a different objective profile — inter-response variability and speech-language measures — confirming a structurally different underlying condition.
Clinical Validation Program
The platform’s diagnostic models were developed and validated across five clinical studies conducted in collaboration with academic medical centers, including the Johns Hopkins Department of Neurology.
873
participants completed 1,378 assessment sessions across the full validation program.
353
community-dwelling adults tracked at baseline and 12-month follow-up in an NIH/NIA-funded primary study.
Gold-standard
MCI classifications benchmarked against neuropsychological reference batteries used in clinical practice.
Diagnostic Groups Studied
Healthy controls · Amnestic MCI · Non-amnestic MCI · Late-life depression · Comorbid MCI + LLD · Parkinsonian disorders
Diagnostic Performance & Subtype Accuracy
A multi-domain risk score derived from the platform’s cognitive factor scores separated each clinical group from healthy controls with the following accuracy, compared to the Montreal Cognitive Assessment (MoCA), the most widely used cognitive screen in clinical practice.
Clinical cohort
Platform AUROC
Standard cognitive screen
Non-Amnestic MCI
0.97
0.70–0.80 general MCI detection only
Parkinsonian Disorders
0.94
Cannot differentiate subtypes
Amnestic MCI
0.94
0.70–0.80 general MCI detection only
Late-Life Depression
0.92
Cannot differentiate from MCI
Mildly Affected MCI
0.89
Fails to detect earliest presentations
Comorbid MCI + LLD
0.73
Confounded by overlapping presentations
Cross-study generalizability: AUROC 0.85 for very mildly affected community-dwelling adults.
A diagnostic model trained on one participant cohort successfully detected the MCI31 group at an AUROC of 0.85 (95% CI 0.72–0.97), demonstrating transfer across recruitment sources and disease severity levels.
Longitudinal Monitoring & Treatment Safety
Traditional cognitive screens are subject to a well-documented confound: practice effects. Patients perform better on retest simply due to familiarity with the test prompts — not clinical improvement. This makes them unreliable instruments for longitudinal tracking.
Elimination of practice effects
Whitney Health’s architecture produces equivalent, parallel assessment forms across repeated administrations, eliminating score inflation from traditional instruments. Longitudinal changes reflect true clinical change.
Sensitivity to treatment response
Supplementary validation demonstrated statistically meaningful signal separation between baseline and post-treatment assessments in patients undergoing ketamine therapy, confirming sensitivity to real treatment-driven functional change.
Anti-amyloid monitoring
Validated sensitivity to pre-to-post functional change translates directly to anti-amyloid therapy monitoring: tracking cognitive endpoints and detecting safety signals, including early ARIA-related functional changes, non-invasively and at scale.
Additional Platform Properties Validated in This Program
Remote, unsupervised administration
Completed without clinical supervision by community-dwelling adults on their own iOS and Android devices across ages 57–94. Completion rates and data quality were sufficient for clinical-grade analysis, validating primary care and telehealth deployment without on-site neuropsychological staff.
Multi-specialty assessment in a single session
The platform assesses the functional equivalent of neuropsychology, psychiatry, neurology, and speech-language pathology domains within a single self-administered session of approximately 60 minutes — confirmed feasible and acceptable without specialist involvement at the point of delivery.
Peer-Reviewed Publications
The following publications document the platform’s clinical development and validation across MCI, stroke, aphasia, late-life depression, and healthy aging.
Sloane KL, Fabian R, Wright A, Saxena S, Kim K, Stein CM, Keser Z, Glenn S, Hillis AE. Supervised, self-administered tablet-based cognitive assessment in neurodegenerative disorders and stroke. Dementia and Geriatric Cognitive Disorders. 2023.
Mefford JA, Zhao Z, Heilier L, Xu M, Zhou G, Mace R, Sloane KL, Sheppard SM, Glenn S. Varied performance of picture description task as a screening tool across MCI subtypes. PLOS Digital Health. 2023.
Sloane KL, Mefford JA, Zhao Z, Xu M, Zhou G, Fabian R, Wright AE, Glenn S. Validation of a mobile, sensor-based neurobehavioral assessment with digital signal processing and machine learning analytics. Cognitive and Behavioral Neurology. 2022.
Berube S, Nonnemacher J, Demsky C, Glenn S, Saxena S, Wright A, Tippett DC, Hillis AE. Stealing cookies in the twenty-first century: Measures of spoken narrative in healthy versus speakers with aphasia. American Journal of Speech-Language Pathology. 2019.
Clinical Validation Benefits Summary
Whitney Health bridges the gap between expanding neuro-therapeutic demands and limited clinical resources. The clinical evidence supports the following conclusions for health system leaders evaluating the platform:
Diagnostic precision where it matters most.
AUROCs of 0.94–0.97 for the primary therapeutic targets — substantially above standard-of-care cognitive screens — ensuring eligible patients are identified and ineligible patients are protected.
Deployment without specialist dependency.
A self-administered, unsupervised model validated in community settings across adults aged 57–94. The platform can be deployed across primary care and telehealth lines without adding neuropsychological staff at the point of delivery.
A permanent, auditable clinical record.
Every session produces a structured digital record of cognitive and functional health, supporting diagnostic documentation, HCC risk-adjustment coding, longitudinal care monitoring, and payer negotiation.
Validated in the population your system serves.
Community-dwelling older adults, not controlled research environments. The evidence reflects real-world deployment conditions.
Whitney Health is a product of Kainoa Labs, Inc.