Feature Extraction Using Hybrid Approach of VGG19 and GLCM For Optimized Brain Tumor Classification

A brain tumor is among the illnesses that are fatal. This is the rationale behind the significance of early disease detection. Intelligent techniques are always needed to assist researchers and medical professionals in diagnosing tumors. Today's doctors employ a variety of approaches to identify the illness. The most popular technique involves getting an MRI of the brain and analyzing it to look for specific diseases. However, manually evaluating the MRI pictures is quite complex and time-consuming. As a result, attempts are made to discover novel methods for cutting down on the prediction time. Deep learning algorithms assist researchers in spotting brain tumor. Many deep learning methods are employed, including CNN, RNN, LSTM, and others. There are benefits and drawbacks related to these methods. One of the most widely utilized methods for categorization is CNN. It's critical to identify the best features while classifying the tumor. Resnet, AlexNet, VGGNet, and DenseNet are some of the feature extraction methods employed. In this research, we proposed a method that extracts unique and high-quality features using a hybrid approach of VGG19 and GLCM. CNN is then used to classify the resulting images. The suggested method's performance evaluation metrics—specificity, sensitivity, ROC, accuracy, and loss—are examined. The method yields a 0.98 accuracy. The algorithm's sensitivity and specificity are 0.97 and 0.99, respectively. The performance of the suggested model is examined by contrasting it with the methods currently in use.