MONITORING BRAIN HEALTH

THROUGH SPEECH

Our voice-based technology provides an objective & replicable tool to monitor the evolution of psychiatric disorders, in-clinic or at home.

WE IMPROVE PATIENT MONITORING IN PSYCHIATRY

The Problem

Psychiatry is a major public health issue, with huge (yet increasing) prevalence in a context of restrained public health budgets. Quality of care can also be improved: 1 in 2 patients is re-hospitalized within 12 months; bipolar disorders are diagnosed on average 10 years after the onset of the symptoms; some patients need 2.5 years to find the right antidepressant treatment.

What we do

Callyope enables psychiatrists to deliver personalized and outcome-based care with remote patient monitoring capabilities. We ambition to improve quality of care by shortening the time to find the right treatment, adjust treatment as needed and prevent relapses to avoid rehospitalization.

How it works

The patient performs a 1-minute voice test on a smartphone, either in-clinic or at-home. Callyope’s technology analyzes the patient’s speech to predict psychiatric clinical scores, assess fatigue and estimate cognition deficits.

60 seconds
60 seconds

of voice to assess mental health conditions

baseline
No baseline required

Models work for unknown speakers

Why We Care

1 in 5

Europeans suffer from mental health disorders

4% of EU GDP

Overall costs of mental health (direct & indirect)

2.5 years

Before some patients find the right antidepressant treatment

4x

Premature death rate for patients with severe psychiatric disorders

Who we are

Founders

Martin
Denais

CEO

Rachid
Riad

CSO/CTO

Xuan-Nga
Cao

COO

Scientific & Clinical Advisors

Julie
Grezes

Senior researcher INSERM

Emmanuel
Dupoux

Head of CoML (ENS) & Research Scientist at META

Alexis
Bourla

Psychiatrist - Hôpital Saint-Antoine, ICRIN (ICM), INICEA

Philippe
Domenech

Psychiatrist - Hôpital Saint-Anne & Institut du Cerveau

Supported by

The Science

Our technology examines both the voice (acoustic) and the language (content) of patients’ productions.

We develop our own natural language processing (NLP) and signal processing techniques to automate the extraction of speech biomarkers.

From a 1-minute speech snippet, we extract various features such as vocabulary, speech rate, prosody, intelligibility and perseveration to assess motor, cognitive and psychiatric symptoms.

Select Publications

Predicting clinical scores in Huntington's disease: a lightweight speech test. (2022)

Riad, R., Lunven, M., Titeux, H., Cao, X.N., Bagnou, H.J., Lemoine, L., Montillot, J., Sliwinski, A., Youssov, K., Cleret de Langavant, L., Dupoux, E. & Bachoud-Lévi, A.C. ; Journal of Neurology, 269, 5008-5021.

Read
Emotion expression through spoken language in Huntington disease. (2022)

Gallezot, C., Riad, R., Titeux, H., Lemoine, L., Montillot, J., Sliwinski, A., Bagnou Hamet, J., Cao, X.N., Youssov, K., Dupoux, E. & Bachoud-Lévi, A.C. ; Cortex, 155, 150-161.

Read
A comparison study on patient-psychologist voice diarization. (2022)

Riad, R., Titeux, H., Lemoine, L., Montillot, J., Sliwinski, A., Bagnou Hamet, J., Cao, X.N., Bachoud-Lévi, A.C. & Dupoux, E. ; In Ninth Workshop on Speech and Language Processing for Assistive Technologies (SLPAT-2022), pages 30–36, Dublin, Ireland. Association for Computational Linguistics.

Read
Identification of Primary and Collateral Tracks in Stuttered Speech. (2020)

Riad, R., Rudzicz, F., Bachoud-Lévi, A.C., & Dupoux, E. ; In LREC, (pp 1681--1688).

Read

Contact Us

For more information on how Callyope can help, contact us using this form below: