Revolutionizing Brain Tumor Diagnosis with Enhanced Deep Learning and Transfer Learning Algorithms

Brain tumors are one of the main causes of cancer deaths in the worldwide. The successful treatment of brain tumors and improved patient outcomes depends on early detection and treatment. The manual analysis of MRI scans is time consuming and prone to errors. Deep learning is developed as a most significant tool for automated brain tumor detection. This study analyzes the deep learning algorithms such as convolutional neural network and transfer learning algorithms such as ResNet50, VGG16, EfficientN etBO, and EfficientN etB7 for classifying brain tumor using MRI scans. The Proposed approach involves three stages such as Image pre-processing, Image segmentation and classification stages. Image pre-processing techniques are applied to enhance the visual quality of MRI images like increasing brightness, sharpening, contrast adjustment, noise reduction. while segmentation techniques involves thresholding, canny edge detection, region based watershed segmentation are used to isolate the tumor regions. Subsequently, deep learning models are utilized to extract the features and classifying the types of brain tumor cells. The accuracy of testing the image dataset is CNN achieves 88%, ResNet50 achieves 89%, VGG16 achieves 92%, Efficient NetB7 achieves 93% and EfficientNetBO achieves 95% of the highest accuracy in detecting and classifying brain tumors based on comparative analysis of all the model. The model Performance will be measured using various metrics such as precision, recall, Fl-score, and accuracy. These results demonstrate the potential of deep learning for improving brain tumor diagnosis.