AI Can Pick up Infant Neurological Problems from Intensive Care Video

0
31
AI Can Pick up Infant Neurological Problems from Intensive Care Video


AI Can Pick up Infant Neurological Problems from Intensive Care Video
Credit: andresr/Getty Images

Researchers led by the Icahn School of Medicine at Mount Sinai have developed an artificial intelligence (AI)-based program that can accurately pick up neurological problems in newborns admitted to the neonatal intensive care unit (NICU) from video.

If these findings, published in eClinicalMedicine, are validated and scaled up, then it could provide a valuable additional source of monitoring for these vulnerable patients.

“Although many neonatal intensive care units contain video cameras, to date they do not apply deep learning to monitor patients,” said lead author Felix Richter, MD, PhD, a clinician and researcher at Icahn School of Medicine at Mount Sinai, New York, in a press statement.

“Our study shows that applying an AI algorithm to cameras that continuously monitor infants in the NICU is an effective way to detect neurologic changes early, potentially allowing for faster interventions and better outcomes.”

Richter and colleagues wanted to test if a combination of computer vision technology and “pose AI,” which allows monitoring and assessment of continuous human movements, could accurately assess potential health problems in infants in the NICU using video.

The research team built a large dataset to train their AI comprising 282,301 video minutes from 115 diverse infants. As the infants get older, movements increase but both sedation and cerebral dysfunction reduce movements of infants in the NICU and the AI was able to accurately predict both of these outcomes after correcting for age.

Receiver operating characteristic area under the curves (ROC-AUC) is a way of assessing accuracy of predictions and the scores for the AI for predicting sedation in different circumstances ranged from 0.87–0.91. Similarly, scores for predicting cerebral dysfunction ranged from 0.76–0.91. The team noted that the accuracy of the prediction was not greatly impacted by lighting conditions or angles of camera.

“It’s important to note that this approach does not replace the physician and nursing assessments that are critical in the NICU. Rather, it augments these by providing a continuous readout that can then be acted on in a given clinical context,” explained Richter.

“We envision a future system where cameras continuously monitor infants in the NICU, with AI providing a neuro-telemetry strip similar to heart rate or respiratory monitoring, with alert for changes in sedation levels or cerebral dysfunction. Clinicians could review videos and AI-generated insights when needed, offering an intuitive and easily interpretable tool for bedside care.”



Source link