A decision tree approach for enhancing real-time response in exigent healthcare unit using edge computing

• Published in: Measurement. Sensors Vol. 32; p. 100979
• Main Authors: Siddiqui, Eram Fatima, Ahmed, Tasneem, Nayak, Sandeep Kumar
• Format: Journal Article
• Language: English
• Published: Elsevier Ltd 01.04.2024; Elsevier
• Subjects: Bio sensors; Decision tree; Machine learning; Medical health records; Mobile edge computing; Bio sensors; Medical health records; Decision tree; Machine learning; Mobile edge computing
• ISSN: 2665-9174; 2665-9174
• DOI: 10.1016/j.measen.2023.100979

The aim of today's healthcare services is to provide high quality and real-time facilities and treatment options for their patients and give a patient-centric experience with full support. IoT-Based Healthcare System have improved the quality of healthcare services by enhancing its diagnosis and decision-making accuracy. On the basis of data collected from different medical Bio Sensors and Machine Learning techniques, a patient mortality and treatment can be improved with the help of current medical condition and historical Medical Health Records. In the paper a Decision Tree method has been proposed which will firstly acquire real-time medical parameter-based data from the patient through multiple BS. This data will be fed into the already trained Decision Trees in order to classify the patient into Low Risk/Normal/High Risk Category. Mobile Edge Computing technology is used in collaboration with BS in order to provide ultra-latent computation of BS-generated data and transform it into real-time decision. After severity categorization of the patient, a definite task offloading decision, whether to go for no offloading/Edge Offloading/Collaborative Edge Offloading mode will be taken. This will be done in order to facilitate severe patient with prompt treatment in case of any exigency. The proposed method outperformed Energy-Efficient Internet of Medical Things to Fog Interoperability of Task Scheduling, Optimized Latency Fog Computing and Intelligent Multimedia Data Segregation methods with a total of 88 % of improved system's performance.