Heart failure: AI analyzes patients' voices to understand when they need help.

Monitoring heart failure patients directly at home, using a system that uses Artificial Intelligence (AI) to recognize changes in voice or speech that may indicate the worsening of the disease. This is the goal of Avatar-Sc, the home telemonitoring system developed by Niguarda Hospital, the Angelo De Gasperis ETS Cardiothoracovascular Foundation, the University of Insubria, and the Polytechnic University of Milan.

Symptoms of heart failure

According to the Italian National Institute of Health (ISS), heart failure (or cardiac insufficiency) is the leading reason for hospitalization among those over 65. The underlying causes can vary, from left ventricular systolic dysfunction to damaged heart valves, resulting in the heart being unable to contract and pump blood optimally. In most cases, the disease becomes chronic, meaning there is no complete cure, and treatments are generally aimed at managing the resulting symptoms. These symptoms can vary from person to person, ranging from shortness of breath, swelling of the lower limbs, involuntary weight loss, and a persistent cough.

The effects on the voice

Insufficient contraction of the heart muscle leads to fluid accumulation in the lungs (pulmonary edema), which, combined with the resulting difficulty breathing, can alter the patient's voice. Several research groups are focusing on this latter aspect, attempting to develop noninvasive methods to identify early warning signs of the disease's worsening, simply by monitoring patients' voices.

How Avatar-Sc Works

Avatar-Sc has precisely this goal. The project aims to develop a home telemonitoring system designed to analyze the voice and language of heart failure patients, anticipating patient care, explains Mirco Pezzoli of the Department of Electronics, Information, and Bioengineering at the Polytechnic University of Milan. "Analyzing patients' voices to understand, before it happens, when intervention is needed: this is the challenge we are addressing with Avatar-Sc. Our experience will enable us to develop predictive tools that could be truly useful in daily clinical practice."

Alongside its clinical monitoring objective—not exclusive, but in addition to traditional methods, its creators explain —Avatac-Sc's goal is also to bring patients closer to technology, concludes Tamara Rabà of the Clinical Psychology Service at Niguarda Hospital: "We will work to ensure that the interface not only collects useful clinical data but also supports patients' psychological and relational well-being, enhancing empathy and understanding as treatment tools."