DT-AXYOLOV5: An Efficient Digital Twin–Assisted Deep-Learning-Based Blockchain Framework for Patient Discomfort Detection in Smart Healthcare System

Advancements in IT technologies have garnered vast numbers of data on healthcare activities, which helps doctors to diagnose diseases at low costs. Digital twin (DT) is a technology that imitates a physical entity to generate digital data representations. This technology becomes more powerful when it is combined with emerging technologies such as the Internet of Things, machine learning, and blockchain to monitor patients’ health. The proposed system contains a step-by-step process for recognizing patients’ diseases in the hospital through discomfort detection. First, the Azure Digital Twins Python package collects the video input of patients from the hospital by using an Internet protocol (IP) camera, which is preprocessed for machine-learning model AX-YOLOV5, using the AlphaPose library to recognize the 18 key points of human organs. The key points are used to identify the body position of a patient, either lying on a bed or sitting. The temporal thresholding technique recognizes health issues by how repeatedly the coordinates of the key points of the human body move within a certain period. Moreover, the coordinates of the key points are assessed for identifying the correct disease. Additionally, the blockchain-based practical Byzantine fault tolerance (pBFT) algorithm effectively stores and protects individuals’ healthcare data. Finally, the efficiency of the proposed system uses calculations that are based on detecting patient discomfort, model training and testing, latency, and data-processing cost. According to the experimental results, the proposed system’s efficacy rate for recognizing the disease can reach 98.3%.