Prediction of clinical risk factors in pregnancy using optimized neural network scheme

Abstract

Women should be aware of prenancy related health issues. A user-friendly model is developed in which the patients can use as well as clinicians to determine the risks associated with foetal development inside the womb, birth weight, whose effects are typically linked to the mother through biological relationships. Recent advances in computer vision and artificial intelligence offer new techniques for automated evaluation of medical images across a variety of fields, including ultrasound (US) images. Enhancing the detection of the estimated foetal weight (EFW) and mother-foetal disease computations can aid obstetricians in making decisions and reduce perinatal issues. This study aims to build a birth weight classification and prediction of relevant parameters during delivery. In this data analysis suite, exploratory data analysis is performed as part of the data pre-processing to investigate the fundamental information and transformational properties. For feature extracting model, the Advanced Dynamic based Feature Selection (ADFS) algorithm has been used which is optimized using the enriched elephant herding optimization algorithm (EEHOA). The multiple feature estimation is classified using augmented recurrent neural network classifier (AURNN). The findings of analyses with graphical representations have been interpreted through the application of visual analytical techniques.