Samsung has announced that its Galaxy Watch may be able to predict a fainting episode or blackout before it happens, according to a joint clinical study conducted with Chung-Ang University Gwangmyeong Hospital in Korea. The study used the device's photoplethysmography (PPG) sensor to analyze heart rate variability data and then applied an AI algorithm to predict VVS during head-up tilt testing. The AI model predicted faints episodes up to five minutes before they happened with 84.6% accuracy and 90% sensitivity. However, there are significant limitations, including the 64% specificity of the model's accuracy in a controlled tilt-table lab and potential for significant alerts generated when a person is not about to faint.