Predicting vital sign deterioration with artificial intelligence or machine learning

Excerpt

Cardiorespiratory instability of patients during their hospital stay is a frequently occurring, undesired complication that often requires prompt treatment to prevent the downstream consequences of reduced oxygen delivery to tissues. This is one of the primary reasons why such patients have their vital signs, such as heart rate (HR), mean arterial pressure (MAP), respiratory rate, and peripheral oxygen saturation (SpO₂) more or less continuously monitored, at least in high care units such as the operating room, the intensive care unit, and the post anesthesia care unit. In this way, real-time detection of e.g. tachycardia, hypotension or hypoxia has made it possible to promptly react to deterioration, i.e. treating it shortly after it has occurred. …