Indonesian Doctor Develops AI to Help Detect Early Heart Failure

An Indonesian cardiologist has developed an artificial intelligence (AI) based technology that has the potential to help detect the risk of deterioration in heart failure patients early. The technology, called the Novel Auscultation Device of Artificial Intelligence for Heart Failure (NAVI-HF), is expected to serve as a tool for medical staff in determining treatment steps for patients before they are discharged from hospital. NAVI-HF was developed by Dr. dr. Rony M. Santoso, Sp.JP, Subsp. K.I.(K), FIHA, a cardiologist at Primaya Hospital Tangerang. The development of this technology is part of his doctoral research at the Faculty of Medicine, University of Indonesia (FKUI). Unlike conventional stethoscopes that rely on a doctor’s hearing ability, NAVI-HF utilises an AI algorithm to analyse patients’ lung sounds. The technology is designed to detect signs of fluid accumulation or pulmonary congestion, a condition that is often difficult to recognise through standard physical examinations. Pulmonary congestion is one of the main causes of rehospitalisation for heart failure patients after they are discharged. This condition often does not present clear symptoms, risking late detection. With the help of AI, medical staff are expected to identify patients at high risk of deterioration more quickly. The working mechanism of NAVI-HF is relatively simple. The device records chest sounds from five examination points for about one minute. The recording is then processed using an AI algorithm capable of recognising specific sound patterns as indicators of pulmonary congestion. The results provide additional information for doctors to determine the patient’s condition. In a study involving 246 acute heart failure patients, NAVI-HF showed promising results. The technology demonstrated an accuracy rate of approximately 86 percent, sensitivity of 91 percent, and specificity of 82 percent compared to lung ultrasound. A six-month follow-up study indicated that patients with a positive NAVI-HF result had an approximately 1.6 times higher risk of rehospitalisation due to heart failure than those with a negative result. However, the developer emphasised that NAVI-HF is not designed to replace a doctor’s role in establishing a diagnosis. The technology serves as a tool to accelerate the identification process of high-risk patients. In the future, NAVI-HF has the potential to support telemedicine services and home-based monitoring to improve the quality of heart failure management in Indonesia.