Applying K-Nearest Neighbors (Kim) Classifier For The Prediction Of Carotid Intima-Media Thickness

The ability to predict the progress of silent disorders that yields to high-risk disease prevention is a key success in health services. As atherosclerosis runs silently inside our arteries, the ability of predicting its existence noninvasively may contribute significantly towards heart attack and brain stroke pre-detection and prevention. This study developed a KNN-based classifier for predicting the high-risk atherosclerosis based on the analysis of photoplethysmogram waveform. The developed model showed an overall accuracy of 85.185%, 73.58% specificity, and 90.8% sensitivity. The obtained results strengthen the ability of KNN to classify patients based on their atherosclerosis progress into highrisk patient or normal patient. This model can be used to assist the evaluation of the silent progress of atherosclerosis, arteriosclerosis, arterial stiffness, heart attack, and brain strokes in clinical settings using the non-invasive, affordable, and easy to implement PPG technique.