SoundWell: AI Speech Biopsy Platform
Overview
SoundWell is an AI-powered speech biopsy platform designed to transform the human voice into a clinically actionable biomarker for screening and monitoring cardiac conditions such as congestive heart failure (CHF). The goal is to build a mobile application that leverages speech analysis and artificial intelligence to offer a non-invasive, low-cost, and scalable tool for early detection, ongoing assessment, and timely intervention in CHF.
The Issue
CHF is one of the most prevalent and costly chronic conditions in the US, accounting for 1.2 million annual hospitalizations and approximately 20% readmission rates within one year of discharge. In the U.S., approximately 1 in 4 individuals will develop CHF in their lifetime. There is a critical unmet need for early warning systems that can continuously and passively monitor patient status, both in and out of clinical settings, and support timely adjustments to care plans.
Approach
SoundWell integrates advanced speech pathology and machine learning to detect CHF-related changes through everyday speech. The team’s AI models are trained on acoustic and prosodic features to recognize signs of fluid overload with clinical-grade accuracy. They are developing a mobile app prototype to capture speech data in real-world environments, with the long-term goal of evolving it into a dual-facing tool for both patients and clinicians. Through voice-based monitoring, we aim to reduce avoidable hospitalizations, improve quality of life, and lower healthcare costs.
MTM Student Engagement
The MTM team will contribute across technical, clinical, and business domains. Areas of focus include refining AI models, improving user experience and human factors design, supporting clinical trial development, building a business model, and developing the minimum viable product.
Project Brief
