An Integrated CNN-RNN-SVM Framework for Classification of MRI Brain Tumor Images

The early identification and categorization of brain tumor through MRI scans are pivotal for effective medical intervention. The present article encompasses a novel integrated framework that combines traditional machine learning and deep learning methods to categorize images of brain tumors. Utilizing the VGG-19 model pre-trained on ImageNet, we extract high-level features from MRI images, which are further processed by a Long Short-Term Memory (LSTM) framework to extract spatial and temporal dependencies within the data. To manage the high-dimensional feature space effectively, we employ Principal Component Analysis (PCA) for dimensionality reduction, followed by a Support Vector Machine (SVM) for the final classification task. We utilized a variety of data augmentation approach to enhance the capability of the architecture to generalize. Additionally, we fine-tuned the training parameters by employing the Adam optimizer along with early stopping and learning rate decay strategies. The model demonstrated exceptional precision, recall, and Fl-score metrics, with an accuracy of 97.86%. This study not only validates the effectiveness of integrating CNNs, RNNs, and SVMs but also opens avenues for future research in medical image analysis using hybrid deep learning frameworks. Experimental outcomes demonstrate that the proposed model significantly improves the accuracy of brain tumor classification compared to previous methods, offering a promising tool for aiding radiologists in the rapid and accurate diagnosis of brain tumors.