Doctor Develops AI Tool to Reduce Heart Failure Patient Readmissions

A doctor is developing an innovative device that uses artificial intelligence to detect residual fluid in the lungs of heart failure patients, aiming to prevent unnecessary hospital readmissions. Dr. Rony Marethianto Santoso, a cardiologist at Primaya Hospital Tangerang, is leading the creation of the Novel Artificial Intelligence-based Heart Failure Device (NAVI-HF).

The portable device combines a digital stethoscope with AI algorithms. It analyzes chest sounds through an application on a smartphone, allowing doctors to more practically, quickly, and objectively detect signs of pulmonary congestion. "If a patient still has fluid in their lungs, do not rush to send them home. That is what we want to detect earlier," said Dr. Santoso.

Dr. Santoso explained that current decisions to discharge heart failure patients often rely on a doctor's clinical judgment and limited examinations, primarily using a stethoscope. This traditional method can be subjective, depending on the doctor's experience, and may miss subclinical congestion, leading to patients being sent home despite a risk of readmission.

While lung ultrasound offers more accuracy, its use is not yet routine due to equipment, expertise, time, and cost constraints. NAVI-HF is designed as a "second layer" in decision-making. The system analyzes lung sounds via the digital stethoscope and alerts if fluid is detected, indicating the patient is not yet safe for discharge and requires further treatment. Initial research on 246 patients showed the algorithm achieved 86% accuracy, 91% sensitivity, and 82% specificity in identifying residual pulmonary congestion.

“

If a patient still has fluid in their lungs, do not rush to send them home. That is what we want to detect earlier.