Science & evidence

Grounded in peer-reviewed science.

Each biomarker stream rests on published research, not internal claims. The platform is co-developed with senior academic neurologists and validated in hospitals.

The biomarker streams

Four signals, each with a published basis.

Voice & speech analysis

Validated across more than ten peer-reviewed studies as a marker of neurocognitive change.

Winterlight Labs; Nature Digital Medicine (2022).

Eye tracking / saccadics

An established Alzheimer's and Parkinson's biomarker in the published literature.

EAN-validated; WHO-approved assessment tools.

Neurocognitive performance

Built on international gold-standard instruments scored for reaction time and accuracy.

MoCA, Trail Making, Stroop; WHO-approved.

Facial expression detection

Validated for early dementia and neurocognitive-disorder diagnostics, captured non-invasively via the smartphone camera.

Published micro and macro-expression literature.

The evidence machine

A 3.000-patient trial, designed for regulatory weight.

The trial design is built so that one study supports the full EU regulatory file. This describes the trial design and is an objective, not a completed result.

COHORTS

Three groups

Healthy controls (45 to 65); MCI / early Alzheimer's risk; early Parkinson's risk.

SITES

Academic centres

We work with several academic institutions.

DURATION

18 + 6 months

Eighteen months of data collection plus a six-month follow-up period.

REGISTRY

Pre-registration

Planned pre-registration on ClinicalTrials.gov; one trial supports the full EU regulatory file.

Partners

Validated with hospitals and academic centres.

Deep Medicine is being developed with the help of several clinical and academic partners. If your organisation is interested in joining, contact us.

Advisory board

Announced soon.

How the technology works