Brain tumors are a significant cause of mortality worldwide, highlighting the need for accurate and efficient diagnostic tools. Magnetic Resonance Imaging (MRI) and Computed Tomography (CT) provide valuable imaging data; however, manual interpretation remains labor-intensive and prone to variability. This study introduces an automated framework for brain tumor detection that integrates image enhancement, segmentation, and classification. Preprocessing is performed using Contrast-Limited Adaptive Histogram Equalization (CLAHE) and diffusion filtering to improve image clarity. Tumor regions are segmented through the Fast Marching Method (FMM), and classification is carried out using an optimized Support Vector Machine (SVM). Evaluation on a Contrast-Enhanced Magnetic Resonance Imaging (CE-MRI) dataset covering gliomas, meningiomas, and pituitary tumors demonstrates strong results, with sensitivity of 0.98, specificity of 0.99, overall accuracy of 98.6%, and a Dice Similarity Coefficient (DSC) of 0.963. The proposed method achieves high performance while reducing processing time to 0.43 s per image, surpassing several existing techniques. These findings indicate that the framework offers a practical and efficient solution for clinical brain tumor diagnosis, with potential for further improvements through integration of multiple classifiers to enhance robustness.
Asiri et al. (Thu,) studied this question.