Internet of medical things for VTE patients in ICU: A self-attention mechanism-based energy-efficient risk identification scheduling algorithm

Abstract

The explosive growth of medical data has dramatically increased the demand for computing power, resulting in insufficient spectrum resources and communication overload. Hospitals need to invest much money to expand computing resources. Various diseases require varying degrees of multi-sensor and continuous monitoring. Take venous thromboembolism (VTE) patients in the intensive care unit (ICU) as an example, enlargement of the right heart, widening of the pulmonary artery, and abnormal results of myocardial enzyme examination maybe lead to sudden death within a short time in the ICU inpatient ward. Steady and dynamic health monitoring is essential. Patients’ immediate risk perception can significantly improve medical efficiency and reduce adverse consequences. How to provide a more efficient and secure full-time monitoring scheme, dynamically adjust the workload, and allocate computing tasks and requests reasonably is a practical problem to be solved urgently. First, this paper defines a task similarity to measure the similarity between different task packages and determine the priority of tasks to avoid forwarding highly similar task packages and reduce energy consumption. Second, the edge gateway caching mechanism with a self-attention mechanism is constructed, which changes the centralized scheduling mode of traditional cloud computing, devolves the coordination function to the edge, and divides the network into multiple local sub-networks. The central node of the sub-network determines the scheduling scheme. The experimental results show that the system can ensure the quality of service and use the edge’s limited computing resources, effectively shield the inefficient data transmission requirements, reduce the use cost and medical quality, and has a specific theoretical and practical value.