Optimized Cascade and Elman Neural Network for CT Image-Based Brain Stroke Classification

The brain is an intricate organ that regulates a variety of bodily activities. Brain stroke happens when the artery that supplies blood to the brain is either blocked or breaks. The main objective is to propose image analysis techniques for identifying the brain stroke lesion. Thus, this provides an efficient solution for detecting and classifying stroke lesions present in CT brain images. This paper developed an Optimized Cascade and Elman Neural Network (OCENN) for the classification of brain stroke from the CT images. Initially, the CT images of brain stroke are gathered from online sources. After that, the noise removal of the image is obtained by using a wiener filter. The brain stroke lesions are segmented by using expectation-maximization (EM). After that, the important features are extracted. Finally, the OCENN is used to classify the brain stroke. To justify the performance of the proposed methodology, it will be implemented and performances will be evaluated. For comparison purposes, it will be compared with the conventional techniques with various parameters of accuracy, precision, recall, sensitivity, specificity, F _Measure, and kappa. The result will show that the proposed methodology will achieve efficient outcomes in system performance and enhance the accuracy of 98.60% in brain stroke detection.