Brain Tumor Segmentation Based on Convolutional Neural Network and Integration of Deep and Shallow Features

Segmentation of Brain tumors refers to one of the most challenging problems in analyzing a medical image. To establish accurate brain tumor region delineation, brain tumor segmentation is used. Feature extraction refers to one of the basic steps in the processing of an image that aids classification. Various features’ kinds are extracted from MRI images. The traditional classifiers of Machine learning need hand-crafted features that are time-consuming and susceptible to errors made by humans. Against this, Deep learning is too powerful in terms of feature extraction and has already been extensively applied for the aims of classification. The presented technique combines multiple feature extraction strategies (integrating deep and shallow features). Thus, shallow methods extract image features manually and based on prior knowledge from the image using image processing methods. Deep approaches use machine learning and neural networks to automatically extract image information. The proposed method of this paper for image segmentation is based on the combination of the PSO evolutionary algorithm with the K-means clustering algorithm. For image processing, the K-means algorithm works perfectly. In the presented technique, PSO is applied to identify the optimum centers of the cluster. K-means algorithm outcome depends on the basic solution and has good convergence. In Google Colab, the proposed model used the Python programming language to diagnose precision, accuracy, sensitivity, error matrix, and receiver operating characteristic (ROC). Outcomes illustrate that the presented model has a high performance in brain tumor diagnosis. That obtains an accuracy of 87.94% and an average precision of 88.35%.