ABSTRACT
eHealth-related recommender systems can be useful for cardiac patients to obtain the most effective advice in the absence of a medical expert. The existing eHealth system provides health maintenance information using classification techniques. Furthermore, existing recommender systems for cardiovascular disease (CVD) provide the same recommendations for all patients of the same category regardless of their age and sex. Study shows that same advice may not suitable for different genders of the same age. As heart size is different in males and females of the same age, the heart rate of females is faster at baseline of electrocardiography compared to the heart rate of males. Long QTc and smaller QRS duration are also observed in females, while males have shorter QTc waves and long QRS duration. Likewise, the same advice may not suitable for different age groups facing the same disease as kids have a faster heart rate than adults. We concluded that classification techniques could not produce accurate medical suggestion for patients facing heart ailment by looking at these facts. The proposed model develops an efficient hybrid recommender system for CVD to give appropriate suggestions to the cardiac patient in cardiologists' absence. The first part of the proposed model collects clinical test results and diagnoses heart disease and classifies it. The second part of the proposed model intends to provide medical advice related to the patient's diet plan and physical activities. Rules are defined to identify the gender and the age of patients; after that recommender system provides the most useful advice for the patient. To analyze the proposed model's performance, the dataset is collected from the coronary care unit (CCU) of a well-known hospital with the help of medical experts and cardiologist. The quantitative and qualitative analysis of the proposed model indicates that it outperforms accuracy compared with state-of-the-art methods.
