Multi-source data fusion framework for remote triaging and prioritization in telemedicine

The more the worldwide population gets older, the bigger is the need for technologies,computerized software algorithm and smart devices to monitor and assist patients anytime anywhere so as to give them a more independent lifestyle. Young people and children can also take advantage of the benefits of being monitored by healthcare systems not only in hospitals but also by using telemedicine when they are at their own homes. Consequently, in order to accommodate the increasing number of users, the remote patient monitoring is one of the issues that telemedicine and Wireless Body Area Networks (WBAN), have to tackle on, and it constitutes the main focus of this research. In order to provide healthcare services for a huge number of users, the healthcare providers triage the patients. Triaging involves an initial sorting of patients who arrive at the Emergency Department (ED) in order to prioritize the most emergency patients and to ensure providing them the appropriate and rapid healthcare services. Triage is a complex decision-making process, and as a result several triage scales have been designed to guide the triage nurse inside ED to a correct decision. This thesis proposes a multi-sources data fusion framework to improve the healthcare scalability efficiency by enhancing the remote triaging and remote prioritization processes for the patients who are in places that are far from the ED and with no triage nurse. The proposed framework is also used to provide users with the compatible healthcare services over telemedicine systems. The proposed framework named Multi Sources Healthcare Architecture (MSHA) considers multi- heterogeneous sources:sensors (ECG, SpO2 and Blood Pressure) and text-based inputs from wireless and pervasive devices of WBAN. As telemedicine consists of three tiers (Sensors/ sources, Base station and Server), the simulation of the proposed frameworks as fusion algorithm in the base station and as healthcare services algorithm in the server was demonstrated. In order to save more lives especially for the users who have the most emergency case, the research main goals were set to be for the achievement of a high level of accuracy in prioritizing and triaging remote patients. The role of multi sources data fusion in the telemedicine healthcare services systems in terms of estimating the patients’ medical status has been demonstrated. Five triage levels, which are different levels of medical emergency (risk, urgent, sick,cold state and normal), were presented in this study. The advantages and the role of each level in term of triaging and prioritizing the remote patients were discussed. The research methodology were presented in four connected phases (preliminary study,modelling, development and evaluation) and each phase has certain goals. In addition to that, the demonstration on how MSHA can be applied in a healthcare telemedicine environment was presented using a proposed scenario. Simulation results, based on datasets for different symptoms of heart diseases,demonstrate the superiority of MSHA algorithms as compared to benchmark algorithms in terms of triaging and prioritizing patients remotely and also in terms of data fusion processing in healthcare application. The accuracy of MSHA in triaging the patients who are in Risk level was 100%, while it was 77% for benchmark improvement method, 84% for two sources fusion and 85% for three sources fusion. Moreover, the MSHA accuracy of triaging patients for Sick level was 100%, while it was (54%, 61% and 63%) for the other algorithms respectively. In addition, the MSHA accuracy for triaging the patients who are in Normal level was 100% and it was (66%,87% and 88%) for the other algorithms respectively.

Preview PDF FK 2015 97IR.pdf Download (1MB) | Preview

Actions (login required)

View Item