Design of Cognitive Healthcare System for Coronary Cardiac Disease Detection

Design of Cognitive Healthcare System for Coronary Cardiac Disease Detection

Mihir Narayan Mohanty (Siksha ‘O' Anusandhan University, India)
Copyright: © 2019 |Pages: 34
DOI: 10.4018/978-1-5225-5793-7.ch001

Abstract

This chapter focuses on clinical decision system (CDS) uses in healthcare units. In this chapter, cognitive approaches are taken using soft computing techniques to design clinical decision systems (CDS) for modern healthcare units. Cognitive computing-based approach is considered. It focuses on cardiac disease detection exclusively by considering its surrounding factors. Fuzzy logic is utilized as one part. The other part includes diabetic detection using deep neural network (DNN) for the automatic identification of the disease. The experiment was done with the Pima Indian dataset. The classification result has been presented in the result section. The decision system in the healthcare unit is a suitable example of a multi-agent system.
Chapter Preview
Top

Introduction

Human being has always been a curious creature, constantly aiming at increasing his knowledge about the world he lives in. Towards achieving new knowledge, the typical process of learning consists thus of following steps: (1) collecting a large amount of observations, (2) extracting relevant information, (3) designing a general model that best explains past and future observations. Machine learning found applications in many fields such as genetics, search engines, natural language processing, computational finance or stock market analysis and computer vision. In the case of medical applications, the interest for learning-based methods seems more recent. Nevertheless, this trend is increasing, and more works containing machine learning as keyword are published each year.

In spite of their effective usefulness, it is not popular clinically that remains a concern to researchers as well as medical practitioners. The objective is to changes in clinical effectiveness related to patient care. The design of the evaluated CDSs relies on either randomized controlled clinical trials (RCTs) or laboratory experiments to determine the performance of the physicians or systems under controlled environment. Lack of elaborate surveys on methodologies that indicates reasons for not using CDSs by clinicians and their practice patterns. It requires the involvement of other medical professionals and computer applications. These are computer based patient records (CPRs), hospital information systems (HISs), ancillary care systems, physician order entry (POE), etc. (Devi, Ramani & Pandian, 2014; Singh, Mohanty & Choudhury, 2013; Mohanty, 2016). Such research seems useful in solving issues related to acceptance or use of CDSs and their relative relevance.

Cognitive informatics (CI) is an area related to cognitive and information sciences. The focus is on processing of human information, computing and computer applications mechanisms as well as processes (Al-Sakran, 2015).In this the CI focuses on understanding of activities and work processes related to human cognition that include the interventional solutions concerned to finding solutions to engineering, computer applications and information technology for better human activities. In the framework of biomedical informatics, CI helps in describing, understanding, and predicting the clinical work activities and its nature that benefits the patients, clinicians as well as lay public. It assists in engineering development and finding computing solutions to boost clinical practice such as efficient decision-support system, patient involvement by providing a tool for timely medication schedule. It also helps public health interventions by providing a suitable mobile application to determine the spread of an epidemic (Ozcift & Gulten, 2011; Chi, Street & Katz, 2010; Su, 2008).

Many attempts have been made to design automatic machines in the field of intelligent and cognitive field. These cognitive machines must know their environments involving similar machines as well as human beings. The machines may vary from self-evidenced practical reasons like computer maintenance expenses to wearable computing in health care which have the cognitive capabilities parallel to the human brain. The aim of this work is to describe the challenges concerned to this new design paradigm which may take into account the systemic problems as well as the design issues. It also includes the teaching of undergraduates in the field of electrical and computer engineering, researchers, etc. Most studies use an experimental or RCT in order to assess system capabilities in a varying clinical environment for better patient care (Silverman et al., 2015). Some of these researches are in the field of CDS whereas none follow a naturalistic design as far as the routine clinical settings concerned to real patients are considered. The studies mainly focus on physicians but not on other clinicians. Further, evaluation of CDS studies is insulated from evaluations of the informatics applications.

Complete Chapter List

Search this Book:
Reset