Heart Disease Diagnosis System with k-Nearest Neighbors Method Using Real Clinical Medical Records

Heart disease is a serious disease that can lead to the death of a patient. Many types of research have been performed related to heart disease, including computer science-based research. Heart disease diagnosis studies, which aim to predict the type of heart disease from which a patient suffers, are popular and refer to some clinical parameters taken from the patient in the hospital. These data are then computed using data mining or machine learning techniques. Some algorithms can be utilized for this purpose, for example, Naïve Bayes, Support Vector Machine, and k-Nearest Neighbor algorithms. This research intends to diagnose heart disease using a sample of patient data. The data were collected from Harapan Kita Hospital, the biggest cardiovascular hospital in Indonesia. The patient clinical parameters selected are taken from complete data available in the hospital's medical records. The k-Nearest Neighbor method gave good results for heart disease diagnosis, and these results can be further used for another purpose, such as telemedicine or a machine-to-machine (M2M) based healthcare system.