Machine Learning Models Enhance Menstrual Recovery Prediction Using Wearable Data
A recent study has developed a framework for predicting menstrual recovery using machine learning models and wearable technology. The research focuses on secondary amenorrhea, a condition affecting reproductive, cardiovascular, and bone health. The study utilizes a synthetic dataset of 5,000 individuals, incorporating physiological features such as heart rate variability, resting heart rate, sleep, physical activity, skin temperature, perceived stress, age, and duration of amenorrhea. Twelve machine learning models were evaluated for their ability to predict menstrual recovery within three months. The models were tested using both baseline and longitudinal data configurations, with ensemble methods like TabPFN, AdaBoost, and XGBoost showing the highest performance. The study highlights the potential of integrating wearable-derived data into health monitoring systems, although real-world clinical validation is still required. 
