PREDICTING DIABETES BASED ON CARDIOVASCULAR AUTONOMIC NEUROPATHY USING CNN WITH KALMAN FILTER

Diabetes is a chronic metabolic disease caused by a high level of sugar in the circulatory system. Diabetic neuropathy is a dangerous nerve-damaging disease caused by diabetes that severely affects cardiovascular heart rate and blood pressure. Autonomic neuropathy is a common long-term complication of diabetes and can be diagnosed with Heart Rate Variability (HRV) calculated from electrocardiogram recordings. As the progression of diabetes leads to higher complications it is very essential to detect earlier for timely treatment of the disease. Most of the commonly used methods for detecting diabetes are invasive in nature. Due to the invasiveness of current commercial glucose meters, more and more patients are suffering from pain and infections. We employed a non-invasive method in which Convolutional neural network (CNN) is combined with Kalman Filter (KF) algorithm to achieve high accuracy in automatically detecting the abnormality. The maximum accuracy obtained from test data is 98.69%.