Remote Diagnosis of Heart Disease Using Telemedicine

In recent years, telemedicine has been a newly emerging field of technology, systems which aid in remote diagnosis have become a need of the hour. With the population increasing at an alarming rate, an efficient system for rapid and efficient diagnosis is imperative. In this paper we have performed a survey analysis of the existing systems of this sort, and we have performed partial implementation of our own proposed system which is a tool for assisted diagnosis of heart disease through telemedicine. Our proposed system, is a setup for remote diagnosis of patients in rural or non-accessible areas, such as isolated military camps or accident sites where specialized diagnosis and treatment is difficult to get. We have proposed a low-cost solution which comprises of a self-designed digital stethoscope which records and amplifies heart sounds to give us the phonocardiogram i.e. the PCG of the patient, and it classifies these sounds as normal or abnormal. The broad and significant highlights to build this system include- diaphragm as primary sensor for data acquisition from the subject; the data processing and its classification is based on Neural Network using Matlab as the IDE. Neural network used for this application utilizes unsupervised learning i.e it classifies data to inherent structure from input data and no pre-categorization is provided to it. Fine-KNN is classification algorithm in Neural Network that separates the data points into various class with respect to each other's similarities. Even classes are clustered based on similarity of neurons in the neighborhood. Further, this data is communicated to a specialist i.e. a cardiologist in this case, in the cityalong with the patient history for further investigation and the expert medical advice of the specialist is relayed back to the patient, or the doctor at the remote location. Our proposed setup also provides features like having live video conversations with medical experts and a unique identification system along with a fingerprint activated security system.