<![CDATA[Brain tumor classification on Magnetic Resonance Imaging (MRI) scans poses a significant challenge in the fields of radiology and medical technology. To enhance diagnostic accuracy, Convolutional Neural Network (CNN) methods have shown great potential. However, the limitation of having an adequate training dataset remains a major obstacle in developing effective models. This study aims to evaluate the performance of CNN models by applying various data augmentation techniques for brain tumor classification and identifying the most effective augmentation techniques. The augmentation techniques tested include image scaling, random rotation, vertical and horizontal flipping, random brightness adjustments, and combinations of these various techniques. The results indicate that the scaling and vertical and horizontal flipping techniques yield the highest average accuracy of 92.97%, with a maximum accuracy of 100% achieved at the 20th epoch using the vertical and horizontal flipping technique. Thus, it is hoped that the findings of this study can be utilized by other researchers in selecting appropriate augmentation techniques for MRI images.]]>
Copyrights © 2024
