Trends in voice characteristics in patients with heart failure
Heart failure (HF) represents a major health and economic challenge worldwide. The expenses associated with HF are primarily driven by hospitalizations, many of which could potentially be prevented. Current self-management programs, such as monitoring weight and blood pressure, fail to reduce hospital admissions due to their low predictive power for decompensation and high adherence requirements. We hypothesize vocal folds are more sensitive to fluid accumulation. Thus, detecting subtle voice changes may enable early identification of acute deterioration, potentially reducing hospitalizations. This pilot study investigates the potential of voice as a digital biomarker to predict health status deterioration in HF patients.
In a two-month longitudinal observational study, we collected voice samples and HF-related quality-of-life questionnaires from stable HF patients. Participants used a study-specific application that we developed, installed on a tablet for use at home throughout the study. Our goal is to explore the correlation between voice characteristics and HF-related quality-of-life. Additionally, we aim to predict health status based on selected voice features and interpret these acoustic features to identify potential health deteriorations.