Voice-based Heart Failure Severity Prediction Engine

a.cure

South Korea

a.cure is a company developing AI models that combine medical science and artificial intelligence to enable early disease prediction and the prevention of emergency situations through voice data–based AI technology. By analyzing physiological signals embedded in everyday human speech using AI-based analysis, a.cure presents a novel, non-invasive approach to assessing health conditions. Its flagship AI model, “Heart to Voice (H2V)”, predicts the presence and severity of heart failure using voice. Results can be generated from approximately one minute of speech, and prior studies have demonstrated an accuracy of up to 90.4%. This research received the Paul Dudley White International Scholar Award at the American Heart Association (AHA) Scientific Sessions 2023, recognizing the technological excellence of the approach within the global academic community. The model is currently being further advanced through additional clinical studies by integrating heart failure clinical indicators to enhance the severity prediction model. Beyond simple detection of heart failure, the technology is being refined to provide a 4-stage severity classification model. Building on its experience in developing Korea’s first Digital Therapeutics (DTx), a.cure possesses strong capabilities in R&D, clinical study design, product launch, and practical commercialization strategy. The company is strengthening technological reliability through a domestic multi-institutional clinical collaboration network, and based on this foundation, it is preparing for gradual expansion into global markets. Going forward, a.cure plans to expand its technology beyond heart failure to respiratory diseases such as COPD and broader cardiovascular disease areas.