Executive Summary : | Public health resources worldwide are focusing on prenatal health surveillance for high-risk pregnant women to decrease perinatal mortality. The pandemic has degraded the immunologic function of pregnant women, necessitating the development of a non-invasive wearable device with fetal movement detection architecture to monitor them at home. The research uses multi-point Inertial Measurement Unit (IMU) sensors for acquiring fetal signals, followed by signal pre-processing and feature classification and recognition modules. The IMU sensors' angle initialization is done with the correcting function, and the energy value of the FM is calculated using the individual IMU's deviation angle. The FM duration is obtained by the initial time point and end time point, and the total energy calculated by the individual IMU is used to find the FM Relative Force (RF). The proposed wearable architecture is validated through comparative analysis, phantom test, and clinical analysis. The efficiency of the proposed framework is evaluated using evaluation metrics such as F1-score, precision, accuracy, and recall. The phantom design FM simulation system includes the pregnancy woman abdomen structure and FM generation using the steward platform. Clinical analysis will be performed on subjects from sri Ramakrishna Multi-speciality Hospital, Coimbatore, Tamilnadu. The results show that the proposed wearable is reliable, easy to wear, and suitable for long-term monitoring of pregnant women at home. Fetal parameters are monitored even during sleep time, and fetal details are updated in the mobile application, accessible to family members and gynaecologists. |