Source-linked AI summary
Early Detection of Breast Cancer using SVM Classifier Technique
Y. Ireaneus Anna Rejani, S. Thamarai Selvi
TL;DR
The paper addresses detecting weak-contrast tumors in mammograms and extracting features to categorize them. It combines enhancement, threshold-based segmentation, feature extraction, and SVM classification, achieving 88.75% sensitivity on 75 mini-MIAS mammographic images.
Problem
The paper addresses detecting tumors with very weak contrast against mammogram backgrounds and extracting features that categorize them.
Method
The method enhances mammograms, segments tumor regions by thresholding, extracts morphological features, and classifies regions using an SVM.
Results
The paper summarizes the developed method as extracting morphological features from tumor regions and classifying them with an SVM.
Takeaways & Limitations
The proposed pipeline provides an approach for categorizing breast tumor regions from enhanced and segmented mammograms.
Abstract
from arXiv · showhide
This paper presents a tumor detection algorithm from mammogram. The proposed system focuses on the solution of two problems. One is how to detect tumors as suspicious regions with a very weak contrast to their background and another is how to extract features which categorize tumors. The tumor detection method follows the scheme of (a) mammogram enhancement. (b) The segmentation of the tumor area. (c) The extraction of features from the segmented tumor area. (d) The use of SVM classifier. The enhancement can be defined as conversion of the image quality to a better and more understandable level. The mammogram enhancement procedure includes filtering, top hat operation, DWT. Then the contrast stretching is used to increase the contrast of the image. The segmentation of mammogram images has been playing an important role to improve the detection and diagnosis of breast cancer. The most common segmentation method used is thresholding. The features are extracted from the segmented breast area. Next stage include, which classifies the regions using the SVM classifier. The method was tested on 75 mammographic images, from the mini-MIAS database. The methodology achieved a sensitivity of 88.75%.
II. 2. METHODS
The method detects mammographic tumors through sequential enhancement, segmentation, feature extraction, and SVM-based classification. Enhancement combines filtering, top-hat processing, wavelet reconstruction, and intensity adjustment, while segmentation uses thresholding and extracts geometric features.
- Methods: The tumor-detection pipeline enhances mammograms, segments tumor regions, extracts features, and classifies the resulting regions.Enhancement and intensity adjustment improve visibility before segmentation and feature extraction.
- Image enhancement: Enhancement applies Gaussian smoothing, morphological top-hat filtering, discrete wavelet decomposition and reconstruction, and intensity adjustment.Top-hat filtering corrects uneven background illumination when the background is dark.
- Segmentation and Feature extraction: Segmentation converts enhanced mammograms to binary images through thresholding and applies Gaussian smoothing again to reduce noise.Thresholding subdivides the image into constituent regions to support breast-cancer detection.
- Segmentation and Feature extraction: Features extracted from segmented images include area, centroid, major-axis length, minor-axis length, and eccentricity.These measurements characterize properties of the segmented mammogram regions.
- SVM classifier: The SVM classifier separates binary classes with a maximum-margin hyperplane, using extracted attributes and support vectors near the class boundary.The method can use kernels including linear, polynomial, RBF, and sigmoid; RBF is described as the most popular choice.
III. 3. RESULT
The proposed method enhances mammograms through filtering, contrast stretching, background elimination, and multiscale DWT reconstruction before threshold-based segmentation.
- Gaussian filtering uses the image’s standard deviation and matrix dimensions, including rows and columns.
- Contrast stretching is applied to the filtered mammogram image.
- A top hat operation eliminates the image background before the output is decomposed into two scales for DWT reconstruction.
- The reconstructed image is segmented using thresholding.
IV. 4. CONCLUSION: · Springer, 1995
The developed method enhances and segments breast tumors using gray-level information, extracts morphological features from each tumor region, and classifies tumors with an SVM classifier.
- IV. 4. CONCLUSION:: The initial step uses gray-level information to enhance the image and segment the breast tumor.
- IV. 4. CONCLUSION:: Image enhancement and tumor segmentation are performed as the method’s first processing stage.
- IV. 4. CONCLUSION:: Morphological features are extracted from each identified tumor region.
- IV. 4. CONCLUSION:: The extracted morphological features are used to categorize the breast tumor.
- IV. 4. CONCLUSION:: The method concludes with classification of tumor regions using an SVM classifier.
- IV. 4. CONCLUSION:: Overall, the method links enhancement, tumor segmentation, morphological feature extraction, and SVM-based classification.