Abstract—Real-time monitoring through wearable and implanted devices is made possible by Wireless Body Area Networks (WBANs), which have emerged as a key component of modern healthcare. These networks provide substantial advantages for patient treatment by enabling ongoing health data collection. The requirement for fast throughput, low packet delay, Packet Delivery Ratio (PDR), and energy efficiency under dynamic network conditions makes creating a Medium Access Control (MAC) protocol crucial. A Q-Learning-Based MAC Protocol (QL-MAC) designed for slot allocation in WBANs is proposed in this paper. QL-MAC improves network performance across important metrics by dynamically optimizing slot allocation using Reinforcement Learning (RL). By adjusting to different network densities and traffic patterns, the protocol guarantees steady gains in communication. QL-MAC Outperforms Adaptive MAC (ADT-MAC), Dynamic Medical Traffic Management MAC (DMTM-MAC), Traffic Aware MAC (TA-MAC), Multi-Constraints MAC (McMAC), and IEEE 802.15.6 MAC protocols. Experimental results show that QL-MAC achieves higher throughput, reduces latency, maintains a better PDR, and has lower energy consumption, even as network density increases. The benefits of QL-MAC make it especially appropriate for applications where reliable communication and energy efficiency are critical, like chronic disease management and remote patient monitoring. This study also reaffirms the role of machine learning in optimizing communication protocols for next-generation healthcare systems. The results highlight the potential of RL-based approaches to address the unique challenges of WBANs, such as dynamic channel conditions and resource contention. QL-MAC ensures dependable and energy-efficient communication by intelligently managing slot allocation, opening the way for advanced healthcare applications.