Brain Tumor Detection and Segmentation using Improved Bat Algorithm with Improved Invasive Weed Optimization

Abstract

Brain tumours are serious diseases that grow in the brain and are made up of a collection of abnormal and unwanted cells. Therefore, the use of magnetic resonance imaging (MRI) for segmentation and early detection of such tumors is more important to save lives. When it comes to finding people with brain tumors, MRI is very effective and has a slightly higher detection rate than other imaging tests. Detection of brain tumors is an important complex issue in medical imaging systems due to their size, appearance, and irregular shape. Detection of brain tumors is a difficult task with medical imaging systems. In order to address the above-said issue and to develop an effective brain tumor detection technique, an improved Bat algorithm with improved Invasive Weed Optimization algorithm has been proposed in this paper. The improved Invasive Weed Optimization (IIWO) approach has been utilized in the proposed IBIIW algorithm along with the improved Bat algorithm (IBA). In the proposed work weight factor which is used for classification has been improved. Extensive experiments has been performed and it suggests that, the use of MR imaging to segment tumors has a significant influence on early detection of brain cancers. In addition to this, with MR images, the deep learning-based technique produced better detection results. The proposed method has outperformed against the baseline methods in terms of accuracy, sensitivity, and specificity with remarkable results giving values of 0.9394, 0.9281, and 0.9165 respectively.