Researchers have used artificial intelligence (AI) to analyze over 10,000 sleep studies, revealing new health risks that were previously undetected. The study, published in Nature, found that AI could identify five distinct patient subtypes with different long-term health risks.
AI Analysis Reveals New Health Risks
The AI analysis was able to identify patients who were at higher risk for serious health issues, including heart failure, heart attack, cognitive impairment, and atrial fibrillation. The study found that patients in the highest-risk group had twice the odds of death over the next five years compared to those in the lowest-risk group.
The researchers used a type of sleep study called a polysomnogram, which collects data on breathing, brain and muscle activity, and other factors during sleep. The AI analysis was able to identify patterns in the data that were not apparent to human researchers.
Implications for Healthcare
The study’s findings have significant implications for healthcare. The use of AI to analyze sleep studies could help doctors to identify patients who are at high risk for serious health issues, allowing for earlier intervention and treatment. The study’s lead author, Dr. Reena Mehra, said that AI gives us the opportunity to move beyond summary measures and learn from the full richness of sleep physiology.
In a separate study, researchers found that using AI to automate oxygen delivery in hospitals could improve patient outcomes. The study, published in JAMA Internal Medicine, found that patients who received automated oxygen therapy spent more time with optimal oxygen levels and required fewer manual adjustments.
Another study found that an experimental non-opioid pain medicine was more effective than the commonly prescribed opioid Vicodin at relieving pain from abdominal surgery. The study, published in The New England Journal of Medicine, found that the non-opioid medicine had a significantly greater analgesic effect than Vicodin.
Original reporting: Appleton, WI News Feed (HLL/CB) — read the source article.
