Cardiac Affliction Detection Using Improved Grasshopper Optimized CNN

The most significant cause of death globally is cardiovascular disease. Whenever an accurate diagnosis is obtained at the earliest possible stage, cardiovascular problems are avoided. Electrocardiogram (ECG) tests are typically used to screen cardiac conditions, and facilitates healthcare process. Due to the dependence and direction of trained medical professionals, traditional methods of evaluating ECG images require more work and time. As a result, automated methods become necessary and image processing is then used to the medical industry. In this research, a neural network-based, effective computational method for determining heart disease is developed. In general, the input ECG images are first pre-processed using Wiener filter, which calculates the mean and variance and then conducts enhanced smoothing in low-variation circumstances. The K-means clustering method is adopted to perform segmentation on pre-processed image in precise manner. The Grey Level Co-occurrence Matrix (GLCM), which chooses the predominant features is utilized to enable accurate diagnosis of conditions in ECG images, then extracts the pertinent features. Finally, for accurate classification, an Improved Grasshopper Optimized Convolutional Neural Network (IGO- CNN) is used. The obtained outcomes demonstrate that the suggested work assists in the establishment of exact outputs and successfully recognizes the presence of heart disease in ECG images.