Researchers from the Department of Computer Science at the University of Copenhagen and the Danish Rigshospitalet have developed an artificial intelligence algorithm that can enhance the diagnosis and treatment of patients with sleep disorders in a short time. According to Aninews, sleep disorder assessments begin by monitoring an individual's sleep using various measurement tools. Specialists review 7-8 hours of the patient's sleep overnight, manually dividing it into 30-second intervals and categorizing it into different sleep stages, such as rapid eye movement (REM) sleep, light sleep, deep sleep, and so on. This task is time-consuming, but the algorithm can perform it in mere seconds.
Paul Ginon, a professor of neurophysiology and head of the Danish Sleep Medicine Centre, stated, "The algorithm has demonstrated that measurements can be taken with a very high degree of safety using machine learning, saving many hours of work and effectively evaluating and diagnosing numerous patients." He added, "In the Greater Copenhagen area alone, over 4,000 polysomnography (PSG) tests are conducted each year on patients with sleep disorders. It takes a doctor between 1.5 to 3 hours to carry out the necessary analyses. Thus, between 6,000 and 12,000 medical hours can be saved through the deployment of the new algorithm." The researchers hope that the algorithm will assist doctors and researchers worldwide in gaining further insights into sleep disorders in the future.
