Semantic Segmentation of Brain Tumor from MRI Images and SVM Classification using GLCM Features

Abstract

Brain tumor is a critical disease caused by the growth of abnormal tissues in the brain, which can severely impact a patient\'s health. Early and accurate detection of brain tumors is essential to improve treatment outcomes and save lives. This project focuses on the semantic segmentation of brain tumors from MRI images, followed by classification using GLCM features and Support Vector Machine (SVM). MRI images are first pre-processed using median filtering and skull stripping to enhance quality and remove non-brain regions. Thresholding and watershed segmentation techniques are then applied to accurately segment the tumor region from the brain. Once segmented, texture features such as contrast, correlation, and entropy are extracted using the Gray-Level Co-occurrence Matrix (GLCM) method in MATLAB. These features are then fed into an SVM classifier to determine whether the tumor is benign or malignant. The proposed system achieves an average classification accuracy of 93.05%, which is higher than many conventional methods. This computer-aided diagnostic (CAD) approach enhances the speed, accuracy, and reliability of brain tumor detection, assisting radiologists in making effective clinical decisions.

Introduction

This project presents a hybrid method for the automatic detection and classification of brain tumors in MRI images using image processing and machine learning techniques.

Methodology:

1. Preprocessing:

   • Noise removal and contrast enhancement using filters and sharpening tools (e.g., imsharpen(), imadjust(), imgaussfilt()).

   • Grayscale images converted to binary using a set threshold.

2. Segmentation Techniques:

   • SOM Clustering and Watershed Algorithm are applied to extract the tumor region.

   • Range filtering and smoothing enhance textural features.

   • GLCM (Gray-Level Co-occurrence Matrix) features (contrast, correlation, entropy, homogeneity) are extracted from the segmented region.

3. Classification:

   • Support Vector Machine (SVM) is used to classify tumors as benign or malignant.

4. Tools:

   • Implemented in MATLAB, leveraging its built-in toolboxes for signal processing, neural networks, etc.

Challenges Addressed:

• Variability in image quality due to acquisition errors.

• Overlapping of organs and poor contrast in MRI images.

• Presence of artifacts and noise that hinder accurate segmentation.

Key Algorithms Used:

• Watershed Algorithm: Simulates water flow over image topology to separate regions based on intensity.

• Genetic Algorithm (GA): Used for complex optimization, mimicking natural evolution.

• DWT (Discrete Wavelet Transform): Used for multi-resolution analysis and image decomposition.

Research Objectives:

• Acquire and preprocess MRI data (e.g., skull stripping).

• Segment tumor using SOM + Watershed methods.

• Extract texture using GLCM.

• Classify tumors using SVM.

• Evaluate performance using accuracy, processing time, and false positive rate.

• Develop a MATLAB-based GUI for clinical use.

Literature Survey Insights:

• Techniques like Gabor filtering, wavelet transforms, and CNNs have proven effective in enhancing medical image analysis.

• Studies support the use of texture analysis and SVM for classification tasks.

• Multi-step, hybrid methods enhance diagnostic reliability.

Conclusion

There are many different kinds of tumours that might be discovered in today\'s world. There is a possibility that the tumours are malignant across the whole brain, or they might be a mass inside the brain. Take into consideration the following scenario: If there is a mass, then the K-means approach is enough to extract it from the brain cells. It is necessary to remove any noise that could be present in the MR image before the K-means approach is carried out. While the tumour is being excised from the MRI scan, the noise-free image is being supplied as an input to the k-means algorithm. This occurs simultaneously. In the next stage, the cancer will be segmented using fuzzy C means in order to properly extract the shape of the malignant tumour. After that, the output of the feature extraction process will be thresholded. In the next stage, an approximation of the reasoning that underlies the calculation of the position and shape of the tumour will be performed. The results of the experiment are compared to the results of other algorithms in order to draw further conclusions. In order to get more precise outcomes, the approach that has been proposed is used. Using 3D slicers in conjunction with MATLAB, it will be feasible in the not-too-distant future to produce a three-dimensional study of the brain.

References

[1] Chen, C. W, Luo, j., parker, k. J., “image segmentation via adaptive k-mean clustering and knowledge-based morphological operations with biomedical applications, ”IEEE transactions on image processing, vol. 7, no, 12,dec. 1998, pp. 1673-1683. [2] Christine, c., Thomas, f., “a study of efficiency and accuracy in the transformation from rgb to cielab color space,” IEEE transactions on image processing, vol. 6, no. 7, july 1997, pp. 1046-1048. [3] Dhawan, a. P., “a review on biomedical image processing and future trends,” computer methods and programs in biomedicine, vol. 31, no.3-4, 1990, pp.141-183. [4] Gonzalez, r. C.; woods, r. E., “digital image processing, 2nd ed.”, prentice-hall, Englewood cliffs, nj, 2002. [5] ng, h. P., ong, h. H., foong, k. W. C., goh, p. S., Nowinski, w. L., “medical image segmentation using kmeans clustering and improved watershed algorithm, ”IEEE southwest symposium on image analysis and interpretation, 2006, pp. 61-65. [6] Tsai, c. S., Chang, c. C., “an improvement to image segment based on human visual system for object-based coding,” fundamental informatics ae, vol. 58, no. 2, 2004,

Copyright

Copyright © 2025 Paluri Deva Aravind , M. Pavani . This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.