Regresar

Applying machine learning to determine development of ecuadorians neonates

Abstract:

The first step in the care of neonates is important because it allows for the avoidance of many pathologies associated with anthropometric parameters. In the last decade, the public health system in Ecuador has increased its physical infrastructure to provide greater coverage of medical services to neonates. However, the necessary hospital equipment to treat neonates is only found in large hospitals in the main provinces of the country-leaving those in difficult-to-access locations unattended. This research proposes to carry out several machine learning models using parameters such as gender, number of assisted consultations, and anthropometric values, to determine a diagnosis for development stage of a neonate. All models were analyzed based on accuracy, specificity, and sensitivity metrics, to determine which has the best performance. Using this model would assist physicians in giving accurate and fast diagnosis, independently of specialized equipment availability.