Brain tumor segmentation based on kernel fuzzy c-means and penguin search optimization algorithm

Brain tumor is the irregular growth of cells in the brain that can develop into malignant or benign tumors. However, the prediction of brain tumors is the most difficult task in the medical field due to the anatomy of the tumor cells. In recent years, advances in deep learning access to medical diagnostic imaging have led to greater accuracy in short time segmenting brain tumors. In this work, a novel approach based on Segmentation-based Kernel Fuzzy C-Mean (SKFCM) with Penguin Search Optimization Algorithm (PeSOA) with an Adaptive Dense Neural Network (ADNN) classifier was implemented. The MRI images are pre-processed using Bright-contrast Dynamic Histogram Equalization (BCDHE) with a weighted median filter and the multi features are extracted with the Modular linear discriminant analysis (MLDA). The Adaptive Dense Neural Network (ADNN) using a unique SKFCM with a machine learning-based Penguin Search Optimization Algorithm was used to segment brain tumors (PeSOA). The effectiveness of the proposed method was estimated based on specificity, accuracy, sensitivity, and tumor area length in vertical and horizontal locations. The proposed approach progresses the overall accuracy of 1.11%, 4.44%, and 6.18% better than CNN, ANN-fuzzy C-means, and R-CNN, respectively.