Disease management: clustering-based disease prediction

Abstract

Abstract:The latest trends in healthcare industry have made enormous data available which is useful for disease prediction. This may be utilised by provider/payer for disease management study (DMS) for judging the present health condition and analysing disease trends. The present approach depends upon objective analysis which is based on binary value of symptoms and health indicators and direct relation between them. We propose a framework based on analytics to diagnose and predict diseases using enhanced genetic clustering using patient history, symptoms, and existing medical condition. New patient's profile is analysed against existing patterns and mapped to a particular cluster based on the patient's medical parameters. After cluster formation, cohesion and adhesion between the clusters is determined, which helps to predict present medical condition and probability for being prone to a disease in future. This framework provides an efficient disease management based on the preliminary diagnosis of the patient. 70 K. Kaushik et al.Keywords: disease management; disease prediction; disease analysis; clustering; healthcare; big data analytics.Reference to this paper should be made as follows: Kaushik, K., Kapoor, D., Varadharajan, V. and Nallusamy, R. (2014) Biographical notes: Kulvaibhav Kaushik working in research stream at Infosys Limited, has a BTech in Computer Science from IIT Roorkee. He has worked on mobile money, merchant transactions, healthcare, big data analytics, cloud computing, and gamification. Much of his research has been carried out over network security, elliptical curve cryptography, and searchable encryption. He has filed patents for methods of secure data storage at cloud and multiuser encryption methodology. He has a few publications in IEEE and others and has delivered keynote address at various national and international conferences. He is also an editor for a mobile payments magazine.Divya Kapoor is a confident healthcare Quality Analyst with more than five years of experience in insurance and provider sector. She is an American health insurance primer Part A and part B certified professional. She has worked on various information technology systems in claim adjudication, revenue cycle, management, and hospital ecosystem.Vijayaraghavan Varadharajan is Senior Research Scientist at Infosys Limited where he is doing research in the field of security assessment, watermarking techniques, network security, and security in converged networks. He has published many papers in international journals and conference proceedings and authored few patents in the area of network security. He has also refereed for many journals and served as a TPC Member of several IEEE and ACM international conferences. He holds a Bachelor of Engineering degree in EC Engineering, MTech in Advanced computing and PhD degree in Information and communication engineering.Rajarathnam Nallusamy is currently the Founder and Managing

Challenges and Opportunities of Big Data in Health Care: A Systematic Review

“…Big data will play a significant role in this transformation [37]. It will allow the information to be delivered to patients directly and empower them to play an active part in their care [5,15,27].…”

Big data analytics for preventive medicine

“…Results showed that k-mean clustering can handle EEG data efficiently in order to detect epileptic seizure. Kaushik et al [170] applied enhanced genetic clustering to diagnose and predict diseases based on patient history, symptoms, and existing medical condition. New patients' profile is analyzed against existing patterns and mapped to a particular cluster based on the patients medical parameters followed by prediction of medical condition and probability of being prone to that disease in future.…”

A semi-supervised privacy-preserving clustering algorithm for healthcare

“…In the 2009 PBC [19], the organization seeks to classify the fiber streamlines of the HDFT dataset into 20-50 fiber tracts consistently within and between brains using unsupervised clustering methods, such as k-means algorithm [23], k-center point algorithm [24] and ISODATA [25]. Following that, in [26], Kulvaibhav et al propose an enhanced genetic clustering method to diagnose and predict diseases using patient's history, symptoms, and existing medical condition. Afterwards, Banu et al [7] utilize Self Organizing Maps based clustering [27] for brain tumor gene expression dataset.…”