Design of Smart Electrocardiography (ECG) Using Modified K-Nearest Neighbor (MKNN)

Heart disease is a disease with a high mortality risk. Errors in diagnosis increase the risk of death. Some efforts to reduce misdiagnosis are developing algorithms for interpretation of ECG signals. This paper presents and discusses decision support system for ECG interpretation, the types of heart disease, electrocardiogram wave, expert system and smart ECG using MKNN (Modified K-Nearest Neighbor) algorithm. MKNN is a method of classification of types of assisted (supervised). In this experiment, we used ECG data from MIT BIH. Because sample is limited, we divide heart disease into 3 classes, that is normal, myocardial infarction and others. The accuracy of MKNN in this study is 71.2% with a value of K = 3.