🤩 AI Monitors Newborns' Neurological Health
🤩 AI Monitors Newborns' Neurological Health
A team of doctors, scientists, and engineers from the Mount Sinai Hospital and School of Medicine in the United States has trained the algorithms to identify signs of neurological dysfunctions based on the intensive care unit's infant video feeds.
First, the new Pose AI tool "watched" more than 4,700 hours of video footage of 115 infants in Mount Sinai Hospital's intensive care unit. Then, the algorithms compared the babies' movements with their EEG data. As a result, the model learned to predict some critical neurological conditions in newborns with an accuracy of up to 91%—these could be signs of seizures, epilepsy, hemorrhage, and other cerebral dysfunctions.
“We envision a future system where cameras continuously monitor infants in the intensive care unit, with AI providing a neuro-telemetry strip similar to heart rate or respiratory monitoring, with alert for changes in status,” explains Felix Richter, a pediatrician at the Mount Sinai Clinic and the study senior author.
A significant advantage of using AI in this area is a continuous monitoring of neurological indicators. Doctors usually conduct examinations every few hours and still do not always notice early signs of dysfunctions, and little patients themselves cannot explain what is bothering them. The AI-powered technology could speed up medical intervention and save the health of many babies.
Researchers from Mount Sinai are already planning to test their technology in other intensive care units, including on adults, applying it to other neurological conditions.
More on the topic:
