Assisted Artificial Intelligence Medical Diagnosis System for Heart Disease

Abstract

In recent years increase of new and effective medical field applications has critical part in research. Artificial Intelligence (AI) Systems has great influence in the growth of these effective medical field applications and tools. One of the major health problems in both established and developing countries is heart disease. Therefore,
diagnosis to regulate the heart disease is very vital, so that appropriate actions can be taken. The Artificial Intelligence System uses input medical data collected from an existing dataset from Kaggle and applies this data on the artificial intelligence application developed that uses data mining algorithm and a basic model on Zambian patients to see if the model will predict correctly. From the dataset collected 80% was used as training data and 20% was used as testing data. The Bayesian data mining algorithm was used for predicting the risk level and probability of heart disease. The system uses medical parameters to predict heart disease in patients and these parameters are age, sex, blood pressure, blood sugar (mg/dl), cholesterol (mm/dl),
heart rate, exercise-induced angina, resting electrocardiogram, oldpeak, ST-slope and chest pain type. The data set collected by the system went through preprocessing which later supervised learning techniques and prediction model was conducted. Results were produced. Based on the results with the prediction accuracy of 90.97%, our results are in the same range as generated by other algorithms like KNN, Random Forest and Decision Tree algorithm.