An Automated Approach For Brain Tumor Detection And Classification In MRI Using SVM
IOSR Journal of VLSI and Signal processing · 2026-01 · 16 (1) · 09-17
DOI: 10.9790/4200-16010917 ↗Kurz gefasst
Fachartikel in der Zeitschrift IOSR Journal of VLSI and Signal processing, 2026-01. Verfasst von Anaswara Viswanath.
Abstract (Originalsprache)
Background
Brain tumors are abnormal and uncontrolled growths within the brain that can disrupt normal brain function and damage surrounding healthy tissues. Early and accurate detection of brain tumors is critical for effective treatment planning and improved patient outcomes. Magnetic Resonance Imaging (MRI) is widely regarded as the most effective imaging modality for visualizing brain structures and identifying tumors. Automated segmentation and classification techniques using machine learning can support radiologists in accurate and timely tumor detection. This study proposes a Fully Automatic Heterogeneous Segmentation using Support Vector Machine for the detection, segmentation, and classification of brain tumors from MRI images. Materials and Methods: In this study, MRI datasets were collected. Preprocessing included grayscale conversion, adaptive contrast enhancement, and skull stripping using a threshold-based method. Tumor regions were segmented using a combination of morphological operations and the Berkeley Wavelet Transform (BWT). A Support Vector Machine (SVM) classifier with a Gaussian kernel was applied to distinguish tumor tissue from normal brain tissues. The system was evaluated using performance metrics including mean, mean square error (MSE), peak signal-to-noise ratio (PSNR), structural similarity index (SSIM), Dice coefficient, sensitivity, specificity, and accuracy.
Results
The method achieved accurate segmentation and classification of tumor and normal brain tissues. The Dice coefficient ranged from 0.79 to 0.90 across MRI slices, while classification accuracy ranged from 96.51% to 98.51%. The method demonstrated high sensitivity and specificity compared to existing techniques such as ANFIS, Back Propagation, DWA-DNN, and K-NN classifiers. Early detection of tumors and precise localization were achieved, although processing time was longer due to the high-resolution images and large dataset.
Conclusion
The proposed approach enables reliable and accurate detection, segmentation, and classification of brain tumors from MRI images. It provides spatially consistent tumor segmentation, supports clinical decisionmaking, and outperforms several existing machine learning techniques. Future work can further improve classification accuracy by exploring hybrid classifiers and advanced feature selection methods
Zitieren
APA
Viswanath, A. (2026). An Automated Approach For Brain Tumor Detection And Classification In MRI Using SVM. IOSR Journal of VLSI and Signal processing, 16(1), 09-17. https://doi.org/10.9790/4200-16010917
Vancouver
Viswanath A. An Automated Approach For Brain Tumor Detection And Classification In MRI Using SVM. IOSR Journal of VLSI and Signal processing. 2026;16(1):09-17. doi:10.9790/4200-16010917
BibTeX
@article{Viswanath2026,
author = {Viswanath, Anaswara},
title = {An Automated Approach For Brain Tumor Detection And Classification In MRI Using SVM},
journal = {IOSR Journal of VLSI and Signal processing},
year = {2026},
volume = {16},
number = {1},
pages = {09--17},
doi = {10.9790/4200-16010917}
}
EUVIMED