Advanced Heart Disease Detection: A Quantum Generative Attention Model for Sensing Networks

The main challenge of predicting heart diseases accurately and securely on real-time data of wearable IoT-based monitoring systems has been a significant problem because of the high computational cost, low forecasting accuracy, and data security issues in the current systems. This paper outlines a robust, innovative system combining hybrid lightweight cryptography and attribute-based encryption to transmit authenticated, protected data. The wearable sensor data undergo secure encryption and transmission to the cloud, and are subsequently normalised using adjusted Min-Max with decimal scaling and statistical column normalisation. Feature learning is performed using a Prompt-Tuned Multi-Task Taxonomic Transformer, and classification is performed with the Efficient Hamiltonian Long-Range Quantum Generative Attention Network, which is optimized with the Elk Herd Optimizer. The Hungarian heart disease data show that experimental analysis achieves high prediction accuracy, greater security, and reduced computational cost, thereby enabling effective real-time healthcare monitoring.