An Innovative Deep Learning Model for IOT-Based Healthcare System for Diseases Prediction

Abstract

IoT healthcare leverages interconnected smart devices and sensors to monitor patient health in real time, enabling remote diagnostics, continuous tracking of vital signs, and timely medical interventions. By collecting and transmitting health data such as heart rate, blood pressure, and glucose levels to cloud-based systems or healthcare providers, IoT enhances the efficiency, accuracy, and responsiveness of modern medical care.The authors introduce Hashing Probabilistic Deep Learning (HPDL) as a secure framework to perform real-time disease predictions and classifications within IoT healthcare systems. HPDL combines weighted data hashing with probabilistic deep learning to achieve security and accuracy in the processing of health sensor data for disease detection purposes. Real-time transformation of IoT healthcare dataset information which included heart rate, blood pressure, temperature, glucose, oxygen level, and ECG was implemented using weighted techniques alongside hashing. Through the implemented model healthcare professionals gained significant success in diagnosing diseases because it predicted normal ECG detection with 0.95, fever with 0.90 and respiratory issues with 0.88 accuracy. Out of critical cases the system successfully activated alerts in 83.3% which shows its effectiveness in making early medical diagnoses. The proposed approach succeeded in preserving a model confidence rate higher than 85% for each condition that received positive classification while proving its strong reliability.