ABSTRACT: In my presentation, I will explore the area of vocal biomarkers and their impact on clinical research. Vocal biomarkers rely on changes in our voice, caused by different health conditions, to help screen and monitor patients. At the Deep Digital Phenotyping Research Unit, my team has been using AI methods and audio signal processing to study voices for health insights. I will present some key findings and examples of how these biomarkers work.
I will also explain how vocal biomarkers can make clinical research more digital and focused on patients' needs while reducing the effort needed from participants. I will also discuss several challenges we need to address. These include the need for better quality data, overcoming inconsistencies in AI research, avoiding biases in the data used for training AI models and figuring out the best ways to use these technologies in actual healthcare and research settings. I will discuss these issues and suggest ways to tackle them, aiming to improve healthcare through personalized and effective methods.
BIO: Guy Fagherazzi is the Director of the Department of Precision Health (DOPH) at the Luxembourg Institute of Health (LIH), which is composed of 11 units dedicated to epidemiology, biomarker, and clinical research.
He is the Head of the Deep Digital Phenotyping Research Unit, a multidisciplinary research lab at LIH where they 1) develop AI-based digital and vocal biomarkers for remote patient monitoring and 2) conduct data-driven digital phenotyping research using large cohort studies and online data to improve the understanding of the impact of various chronic diseases (diabetes, mental health, cancer, Long COVID...) on the daily lives of patients and populations.
Google Scholar Profile: https://scholar.google.com/citations?user=EKz0WDUAAAAJ&hl=fr