Prenatal screening is essential for identifying any issues early and ensuring maternal and fetal well-being. The majority of people live in rural areas, so they do not receive regular prenatal care at the start of their pregnancy. There has been an increase in the number of unanticipated deaths during delivery. In this paper, a novel fetal health care via sensor-fused IoT (FEST) for a pregnancy woman monitoring system. The device can monitor maternal and fetal health indices such as body temperature, fetal heart rate, blood pressure, breathing rate, fetal movement, PPG, and abdominal ECG while at home. Different sensors, including NTC, piezoelectric, and 3-axis acceleration, are used to identify the issue. Using the DWT, all sensor signals are pre-processed to remove unwanted noise. Using Efficient Net to classify the signals after identification produces the best classification results, regardless of whether the maternal health is healthy or unhealthy. Finally, if the pregnant woman is unhealthy, an alert message is sent to the patient mobile. The simulation analysis of the proposed method achieves an accuracy of 99.58%. Additionally, the proposed model achieves 98.37% overall precision, 96.14% specificity, and 93.98% recall. Comparison of the proposed Efficient Net such as ResNet, Dense-Net, and AlexNet. The Efficient Net achieves a higher accuracy rate than the currently used models. The FEST method approach improves the overall accuracy by 1.58% and 9.27% better than DT-BiLTCN and multi-point IMU sensing, respectively.