eCONSULT+ Project
eCONSULT+: AI and voice biomarkers to transform Parkinson's screening
Diagnosing Parkinson's disease presents a significant challenge in primary care. Early symptoms are often subtle, resulting in delays of several years in diagnosis. This delay is especially significant in areas with limited access to family physicians.
Although teleconsultation facilitates access to care, the lack of a complete physical examination sometimes limits the accurate detection of signs of the disease.
Innovation in diagnostics: vocal biomarkers
Solutions are emerging, including the integration of voice biomarkers analyzed by artificial intelligence (AI) algorithms. These non-invasive technologies offer promising potential to support clinical diagnosis and improve the early detection of Parkinson's disease. Furthermore, voice biomarkers are considered promising for remote healthcare delivery, as they provide complementary information that can help physicians make informed decisions regarding the detection, progression, and treatment of Parkinson's disease.
What is a vocal biomarker? It is a specific "signature" of the voice audio signal associated with a pathology.
Why the voice? The voice is produced by complex neuromuscular mechanisms that can be altered by Parkinson's.
The power of AI: While some changes (intensity, amplitude) are perceptible to the human ear, AI can detect thousands of invisible features through machine learning and deep learning.
Project objective
The project is not limited to testing a technology. It aims to understand how an AI system reconfigures medical practice in primary care.
Our study focuses on three main areas:
Analyze the interactions between the doctor, the patient, and the AI during simulated teleconsultations.
Understand the doctor-AI coupling by examining how data provided by AI supports (or not) clinical reasoning.
Training the professionals of tomorrow: Identifying the skills needed to use AI judiciously and maintain a relationship of trust with patients
A partnership of excellence
This project is being carried out in collaboration with Virtuosis AI , a company specializing in AI-powered speech analysis.
It receives financial support from:
The Montfort Institute of Knowledge
The Research Chair in International Francophonie on Digital Health Technologies .