To solve the constraints of conventional ultrasound diagnosis, the proposed work utilizes artificial intelligence methods for enhancing the prediction of fetal health in early pregnancy diagnosis. The approach uses machine learning (ML) techniques such as Support Vector Machines (SVM), Bagging, Multilayer Perceptron (MLP) together with deep learning models (DL) such as LeNet, GoogleNet and ManualNet for ultrasound examination to analyze health data, thus improving accurate diagnosis, support and timely intervention. Among the ML and DL models, Bagging and LeNet showed the highest accuracy, respectively. The system also provides additional data, including time analysis and data distribution documents, to develop a user application to improve maternal and infant outcomes.