<p>The Internet of Medical Things (IoMT) is transforming healthcare by integrating smart medical sensors for real-time patient monitoring and data-driven diagnostics. However, ensuring the prolonged functionality of these devices remains a critical challenge due to their energy constraints. This study presents a novel hybrid framework that combines machine learning (ML) with reinforcement learning (RL) for intelligent energy management in radio frequency (RF) energy harvesting IoMT systems. The approach employs a predictive ML model to forecast energy consumption based on sensor activity and environmental dynamics, while a double deep Q-network (DDQN)-based RL agent dynamically learns optimal sensor scheduling policies to maximize energy efficiency and data availability. Unlike existing static or heuristic methods, our adaptive control strategy demonstrated superior performance with a cumulative reward of 231.45, an energy harvesting efficiency of 71.8% and a net energy gain of 0.5285 J. These results underscore the framework’s capability to significantly extend system lifetime while maintaining robust healthcare data streams. This work is among the first to integrate DDQN with real-time energy prediction for holistic energy-aware control in RF-powered IoMT, offering a scalable solution for sustainable digital health infrastructure.</p>