The DiaVoc project: Diagnosing vocal characteristics to track patients’ health
This project centers on the diagnosis and monitoring of health conditions that impact patients’ vocal characteristics, including Neurocognitive disorders (NCDs) (signifying cognitive decline), pulmonary disorder (COPD), and heart failure conditions (HF). By utilizing longitudinal voice recordings paired with medical data, we aim to create mathematical vocal characteristics, distance metrics, and machine learning methodologies that are suitable for tracking and categorizing pertinent vocal changes. These characteristics can be helpful for early diagnosis, enhance clinical evaluations of underlying health conditions, boost the effectiveness of treatments, and enhance the prediction of prognosis. A significant strength of this project lies in our transdisciplinary perspectives at the intersection of health science and engineering technology.