Revolutionizing Pregnancy Monitoring: Machine Learning for Trimester Classification Using Biomechanical Data

To ensure the health of both the mother and the growing fetus, pregnancy has to be closely monitored. However, traditional methods of monitoring pregnancy often focus on physiological and biochemical changes instead of biomechanical alterations. Plantar pressure and balance distribution are considered as indicators for monitoring pregnancies, but recent studies mainly involve manual techniques or unsophisticated statistical methods which may not fully capture the complex biomechanical changes that take place within a trimester period. Therefore, this study proposes a trimester-based classification model in pregnant women using various machine learning (ML) algorithms according to plantar pressure and balance distribution data so as to have specific essentials ranges through which gestation development can be monitored effectively. Advanced ML algorithms such as K-Nearest Neighbor (KNN), Random Forest (RF), and Support Vector Machine (SVM) are used in this research to measure subtle kinematics variations. Feature extraction from force plate data is able to capture vital aspects of plantar pressure and balance. First, second, and third trimester samples were employed for constructing the training set and testing the classification model by applying it on them respectively. The method’s aim is to improve prenatal care by giving a digital diagnosis of trimesters based on quantitative biomechanical data. This model aims to prove that machine learning can be implemented using biomechanical data with an accuracy rate up to 80 percent, particularly when KNN approach is used in order to fill the gap in current methods.