A MACHINE LEARNING APPROACH USING STATISTICAL MODELS FOR EARLY DETECTION OF CARDIAC ARREST IN NEWBORN BABIES IN THE CARDIAC INTENSIVE CARE UNIT

A frightening yet common medical emergency is cardiac arrest in infants who have just been delivered. The greatest care and therapy for these infants can only be provided if detected early. Recent studies have aimed to create reliable diagnostic techniques for early diagnosis of neonatal cardiac arrest by finding possible biomarkers and indications of this condition. Echocardiography, computed tomography, and other imaging modalities may aid in the early diagnosis of cardiac arrest. In order to identify cardiac arrest in newborn newborns in the Cardiac Intensive Care Unit (CICU) early on, this project seeks to construct a Cardiac Machine Learning model (CMLM) utilizing statistical methods. Multiple neonate physiological characteristics were used to identify the cardiac arrest occurrences. Predictive models for cardiac arrest were constructed using statistical modeling approaches, including logistic regression and support vector machines. The suggested methodology will be used in the CICU to facilitate the early identification of neonatal cardiac arrest. With a training (Tr) comparison area, the suggested CMLA achieved a delta-p value of 0.912, an FDR value of 0.894, an OR value of 0.076, a prevalence threshold value of 0.859, and a CSI value of 0.842. The suggested CMLA achieved test-specific values of 0.896 for delta-p, 0.878 for FDR, 0.061 for FOR, 0.844 for prevalence threshold, and 0.827 for CSI in a testing (Ts) comparison range. Babies induced cardiac arrest in the neonatal intensive care unit will have a lower risk of death and morbidity as a result.