A Framework for Maternal Physical Activities and Health Monitoring Using Wearable Sensors.

Area of Science:

• Biomedical Engineering
• Wearable Technology
• Maternal Health Monitoring

Background:

• Physical activity during pregnancy significantly impacts maternal and fetal health.
• Continuous monitoring of physical activity intensity is crucial but challenging throughout maternity.
• Existing methods often lack convenience or continuous feedback capabilities.

Purpose of the Study:

• To develop and evaluate a physical activity recognition and monitoring framework using wearable sensors for pregnant individuals.
• To address the challenges of continuous monitoring and activity logging during maternity.
• To provide timely feedback and alerts in unfavorable situations.

Main Methods:

• Utilized a body-worn module with accelerometer, gyroscope, and temperature sensors.
• Transmitted sensor data via Bluetooth Low Energy (BLE) to a Raspberry Pi for feature extraction.
• Employed supervised machine learning classifiers trained on extracted features for physical activity recognition.
• Created a novel dataset of 10 physical activities from 61 subjects across different pregnancy stages.

Main Results:

• Achieved a highest recognition rate of 89% for physical activities.
• Demonstrated the feasibility of using wearable sensors for continuous maternity monitoring.
• Successfully trained machine learning models for accurate activity classification.

Conclusions:

• The proposed framework offers a promising solution for monitoring physical activity in pregnant individuals.
• The system can provide valuable insights and alerts for improved maternal care.
• Wearable sensor technology integrated with machine learning can enhance the safety and well-being during pregnancy.