Obstetrics is a medical specialty focused on pregnancy, childbirth, and postpartum care, ensuring the health of both mother and baby. It involves diagnosing and addressing fetal anomalies during prenatal care. Fetal anomalies, which include structural and functional disorders, significantly impact long term health and are major causes of mortality. This study aimed to address these issues by developing machine learning models for the detection of fetal abnormalities, using explainable AI techniques to clarify model decisions, and creating prototypes for remote diagnosis. We employed a mixed methodology of experimental & design science research, and evaluating twenty one classifiers. From these, three models Tree classifier, XGB classifier, and LGBM classifier were selected and optimized. The LGBM classifier, after hyperparameter tuning, achieved accuracy of 99.53%. To make the models’ decisions transparent, we utilized SHAP and LIME explainable AI techniques, providing reliable insights into the decision making process. Following model development, we created a Django based prototype, focusing on user friendliness and integration into clinical workflows. We conducted usability evaluation on prototype using the SUMI tool, achieving scores between 65.48% and 73.29%, where the maximum possible score is 76. We also conducted a security evaluation by defining a threat model and assessing security risks. In conclusion, our evaluation confirms that the developed models are secure, reliable, and effectively address the research problem of fetal anomaly detection. This work offers a novel, comprehensive solution that integrates advanced machine learning with explainable AI and practical deployment for clinical use. Keyword: Obstetric, Fetal anomalies, Explainable AI, Abnormalities detection, Prenatal care, Security evaluation, Usability evaluation, Prototype.