AI-based visual technology for assessing the neurodevelopment of premature infants

The neurological development of premature infants depends critically on the quality of the first weeks spent in intensive care, yet until now there has been no tool for measuring that quality objectively and continuously. Joint work by HUN-REN SZTAKI and Semmelweis University has produced a camera-based, AI-centred monitoring system that follows the key indicators of neurological maturation contact-free, without burdening the infant, and identifies sleep-wake cycles in real time - established biomarkers of neurodevelopmental outcome. The models were validated on clinical data; sleep-wake detection and posture classification reached reliable accuracy in clinical practice as well. Medically, the technology makes the quality of neurodevelopmental care (NIDCAP, FINE) objectively measurable for the first time where it could previously only be judged subjectively; socially, it can help reduce the risk of cognitive impairment in premature infants and make finite intensive-care capacity go further.