Recent advancements in Cognitive Technical Systems (CTS) have significantly transformed human-computer interaction (HCI) by introducing efficient and natural operating principles. These systems heavily rely on data from multiple sensors, which are integrated using fusion algorithms to enhance their functionality. This study proposes a novel cognitive HCI approach, leveraging body sensor data analytics through machine learning within a Mobile Health Communication D2D cloud framework. The core of this research involves employing a Boltzmann Perceptron Basis Encoder Neural Network to analyze various datasets collected from body sensors within the D2D cloud network. The experimental analysis evaluates the efficacy of this approach across different performance metrics, highlighting its superiority over existing methods. The proposed technique demonstrates enhanced efficiency in processing and interpreting monitored data, contributing to advancements in personalized healthcare and interactive computing environments.