Voice Biomarkers for Dementia Prescreening

See how it works

A scalable prescreening layer

The only large-scale, decentralised, short and self administered cognitive assessment for preclinical dementia screening. Reducing recruitment time and costs, accelerating subject identification and increasing enrollment.

Reducing screening failures with vocal biomarkers

Standard screening

Recruitment pool
PET / CSF testing70% screen failures
Enrolled patients

With Accexible prescreening

Recruitment pool
Accexible prescreeningLow cost · High accuracy
PET / CSF testing40% screen failures
Enrolled patients

Lower costs
Fewer screening failures
More enrolled patients

Voice AI

Clinically validated across 18+ studies and 15,000 participants. AI powered speech analysis to identify potential participants with early cognitive decline and support digital endpoints in trials.

How it works

1
Patient
Takes a short, self-administered voice test.
2
30–60 s of speech
Recorded by phone, website or app.
3
Cloud speech analysis
Accexible's AI models analyse voice and language.
4
Results in under 1 minute
The health professional receives the report.

“When something goes wrong in the brain, it will likely show up in the voice.”

— Camille Noufi.

Preclinical Assessment

Voice biomarkers enable scalable prescreening and earlier identification of individuals who may benefit from new therapies.

Clinical trial prescreening

Voice biomarkers enable scalable prescreening to identify early cognitive decline before costly PET or CSF testing, while maintaining high sensitivity for early cognitive changes.

Recruitment acceleration

Expand recruitment pools through scalable remote screening. Remote voice assessments allow large populations to be screened quickly and cost-effectively, improving recruitment efficiency and reducing screen failure rates.

Longitudinal monitoring

Track cognitive and language changes throughout clinical studies. Speech-based assessments provide a low-cost, repeatable method for monitoring cognitive signals over time with high consistency and minimal participant burden.

Digital endpoints

Complement traditional endpoints with speech-based biomarkers. Accexible’s AI models deliver objective and reproducible digital biomarkers that enhance the precision of cognitive measurement in clinical trials.

Selected partners

Clinical Validation & Regulatory Standards

Accexible combines clinical research, regulatory rigor, and artificial intelligence to develop reliable digital biomarkers for healthcare and clinical trials. Our technology is supported by a growing body of clinical research conducted across multiple countries and diverse populations.

To date, Accexible’s models have been evaluated in more than 18 clinical trials, involving over 15,000 participants. The datasets used to develop and validate our models are multilingual and include research collaborations across Europe and the Americas, enabling robust and generalizable results.

Accexible is built to meet the requirements of healthcare and clinical research environments. The technology is CE-marked and backed by patented IP and developed following ISO 13485 quality management processes, ensuring rigorous development standards for medical technologies. In addition, Accexible adheres to ISO 27001 information security standards, guaranteeing high levels of data protection and privacy.

Advisors

Donald A. Davidoff

Assistant Professor at Harvard Medical School and Chief of Neuropsychology in McLean Hospital

Donald Davidoff, PhD, is chief of the Neuropsychology Department at McLean Hospital and an assistant professor of psychology in the Department of Psychiatry at Harvard Medical School. Responsibilities include management of clinical and administrative staff, clinical care, and training of post-doctoral fellows and psychiatry residents. He founded the Geriatric Neuropsychiatry Unit and for 16 years was the psychologist in charge.

Shwen Gwee

Digital Health and Corporate Innovation Executive / Startup Advisor & Investor

Shwen is a recognized pioneer in digital health and corporate innovation, as well as a well-respected industry thought-leader with almost 20 years of experience in the biopharmaceutical industry, leading digital strategy and innovation across commercial, clinical and corporate functions. He was most recently the VP/Head of Global Digital Strategy at Bristol Myers Squibb (BMS), where he was accountable to the Executive Leadership Team and the Board of Directors. He is an investor in several digital health startups and sits on various advisory boards including Accexible.

Pedro Montejo

Doctor in Medicine specialized in Psychiatry

Dr. Pedro Montejo Carrasco is a doctor in medicine and a specialist in psychiatry. He is the head of the Center for the Prevention of Cognitive Impairment, a center founded in 2006 specialised in early detection and cognitive stimulation in cognitive diagnosis. He is co-author of the UNAM Memory Training Method, which has been the base for several published documents and more than 50 training courses in the field of psychogeriatrics.

With the support of

Program: AI R&D Missions.

Project: The Federated Artificial Intelligence Network Project to accelerate Health Research (TSI-100205-2021-2), has been financed by the Ministry for Digital Transformation and Public Service, through the 2021 Artificial Intelligence R&D Missions Program, within the framework of the Digital Spain 2025 Agenda and the National Artificial Intelligence Strategy, with European funding through the Recovery, Transformation and Resilience Plan.

This project has received funding from CDTI through the aid program for SMEs with the Horizon Europe EIC Accelerator Seal of Excellence.

Program: Assistive Robotics.

Project: Robocat.