Thrombophilia Prediction Using Machine Learning Algorithms

Abstract

Thrombophilia in pregnancy is the result of a complex interaction of inherited and acquired factors, which increase blood coagulation and consequently placental ischemic conditions. Early identification of risk of developing thrombophilia in pregnancy is crucial for implementing preventive measures and personalized therapy. In this study, we propose a novel approach for prediction of thrombophilia in pregnancy utilizing machine learning (ML) algorithms with a particular focus on neural networks. The research is done using a dataset consisting of demographic, lifestyle, and clinical information from a 35 pregnant woman (22 healthy and 13 with thrombophilia). These features are used to train and evaluate different ML models with neural networks and decision trees. The evaluation of the proposed approach involves cross-validation and performance metrics assessment. The results highlight the effectiveness of decision trees and neural networks in accurately predicting thrombophilia in pregnancy risk.