Computational Intelligence and Healthcare Informatics

13Primary Healthcare Model for Remote Area Using Self-Organizing Map Network

Sayan Das* and Jaya Sil

Department of Computer Science and Technology, Indian Institute of Engineering Science and Technology, Shibpur, West Bengal, India

Abstract

Around 66% of total population in India lives in rural areas, where the availability of primary health-care is a real problem due to the lack of physicians and trained manpower. Health kiosks are deployed in rural villages and health workers are engaged in acquiring specific individual health data such as blood pressure, pulse rate, temperature, BMI, and level of oxygen saturation (SpO₂). However, the sensors used to measure the basic health data are not standardized, resulting in measurement errors in symptom values. Therefore, when people are diagnosed based on the measured health attribute values, there is always risk due to limited domain knowledge and skill of the health workers. A fuzzy-based health assistance framework was suggested in this study that analyzes the specific health characteristics of individuals to decide if a person is in good health or needs medical attention. First, to deal with vagueness in measuring error of symptom values, we use fuzzy variables with proper semantics with reference to the corresponding standard values, available in the literature. States of health of the patients are derived by analyzing their symptoms, and it is obvious that instead of an individual symptom, combination of symptoms is more ...