Electrocardiogram Data Capturing System By Using Machine Perception

Abstract

Obstructive Apnea is a breathing based sleeping problem, in which throat tissues drops back in the direction of air passages as well as obstructs the air flow, partly or completely during the rest. Because of lack of air movement in blood, oxygen levels will drop suddenly in boosting high blood pressure as well as strains cardiovascular system, leads to boost the danger of Cardiovascular diseases, Stroke, Excessive Weight, Diabetes Mellitus, High blood pressure, Myocardial infarction (MI) etc. where MI is a cardiovascular disease occurs when the circulation of blood, decreases or quits to a part of the heart, which it brings about heart damage. One of commonly detected method for sleeping conditions is Polysomnography (PSG), which is a lot pricier as well as eats much effort, as a result of these reasons in a lot of the cases sleep problems were undiagnosed. To get over the drawbacks of Polysomnography in an affordable way and in less initiative, existing system uses ECG signals in order to identify OSA. ECG is mostly used to diagnose cardiovascular disease, in order to find MI from ECG where as in traditional standard system assessment of ECG signals via visual analysis can be done by physicians or medical professionals is not effective and time consuming. This paper proposes a system to conquer the disadvantage of standard system, by translating the exact ECG by utilizing machine perception and to identify the problems in ECG, mainly concentrates on OSA as well as Myocardial infraction. Machine perception is the capability of a system in order to interpret the data based on exactly how human beings detects and connects the world around them. It permits the system to collect the information based upon equipment vision with better accuracy as well as to provide it in a form which is more comfortable to the user. Computer vision mostly focuses on acquiring, processing, analyzing, and recognizing images. Here the input information is absorbed kind of images as opposed to signals for accurate ECG interpretation. This can be done by using wavelet changes and Auto Regression (AR) approaches and should able to distinguish between regular ECG and also uncommon ECG and it can be done by using improvised classification Algorithm based on integration of both K-Medoid and also improvised KNN classifier is used to reduce the computation complexity and also enhance the precision by using the hyper tuning parameters.