Closeness factor based clustering algorithm (CFBA) and allied implementations—proposed IoMT perspective

Abstract

Now-a-days, IoMT (i.e. Internet of Medical Things) is a buzzword in the healthcare domain. It can be used as a catalyst for personalized care and living through individual data-driven treatment as well as well optimized devices tailored as per physiological needs of an individual. Conceptualization of IoMT have been seen in different forms viz., mobile and web-based healthcare applications, smart wearable devices, care-kits and to name a few. IoMT in other words will automate healthcare monitoring, enhance its operational efficiency, as these devices will have capability to gather and send large amount of data, dynamically. IoMT devices having capability to collect and analyse real time data will be more preferred. There are few implantable, such devices already exist in practice and this research also suggest one/more way to embed the analysis logic into the device to make it more powerful. IoMT may trigger problems like data-theft, data transfer through insecure connection and irregularities in connections of network. So, by looking at these pros and cons of IoMT—this book chapter discusses variety of formulations of CFBA from IoMT perspective for effectual data analysis. CFBA came into existence in literature in an around 2007 and is evolving continuously in varied forms. Few of them are: probabilistic, Naive bayes based, correlation based, threshold based, distributed form, etc. These different versions of CFBA extended their utility in healthcare in different ways viz. recommending an ice-cream to diabetic patients, personalized diabetes analysis, analysis of pathology reports of diabetic patients, recommending an ice-cream to university grads based on their life style and eating habits etc. Recently, deployment of CFBA on Microsoft Azure public cloud is in progress and this ongoing development is intended to handle diverse chronic disease datasets, to achieve scalability with execution efficiency and primarily to make is domain independent. By considering all these existing and ongoing developments of CFBA, this book chapter enumerates their comparative review, proposes extensibility from IoMT perspective which will in turn be useful for chronic disease data real time analysis, personalized treatments, obtaining recommendation patterns and so on, valuable for all types of healthcare professionals.