Please use this identifier to cite or link to this item:
http://hdl.handle.net/10603/522619
Title: | Mammogram learning system for breast cancer diagnosis using deep learning SVM |
Researcher: | Jayandhi G |
Guide(s): | Leena Jasmine J S |
Keywords: | Computer Science Computer Science Information Systems Deep Learning Architecture Engineering and Technology Softmax Layer Vgg-Svm |
University: | Anna University |
Completed Date: | 2023 |
Abstract: | newline Breast cancer is the most common cancer in women and ranks as the commonest cause of death amongst women in the 35-55 age groups. The cause of this cancer is not understood yet and no curable treatment has been discovered. The objective of this research work is to develop a mammogram learning algorithm to improve the diagnostic accuracy of breast cancer. In this work, an efficient Deep Learning Architecture (DLA) with Support Vector Machine (SVM) is designed for breast cancer diagnosis. It combines the ideas from DLA with SVM for effective classification of mammograms. The state-of-the-art Visual Geometric Group (VGG) architecture is employed in this work as it uses the small size of 3x3 convolution filters that reduces system complexity. The softmax layer in VGG assumes that the training samples belong to exactly only one class, which is not valid in a real situation, such as in medical image diagnosis. To overcome this situation, SVM is employed instead of the softmax layer in VGG. Data augmentation is also employed as DLA usually requires a large number of samples. VGG model with different SVM kernels; Linear (L-SVM), Polynomial (P-SVM), Radial Basis Function (RBF-SVM) and Quadratic (Q-SVM) is built to classify the mammograms. The proposed VGG-SVM system operates under two modes: Local Classification Mode (LCM) and Global Classification Mode (GCM). Though the system classifies the input mammograms into two classes, the former uses the Region Of Interest (ROI) images, and then later uses the whole mammograms as inputs. Also, preprocessing is employed to remove undesirable information to ensure data integrity and improve classification performance. In this stage, morphological operations are employed to remove the information embedded in iv iv the X-ray image and contrast limited adaptive histogram equalization is chosen as a contrast enhancement technique. |
Pagination: | xvi, 120 p. |
URI: | http://hdl.handle.net/10603/522619 |
Appears in Departments: | Faculty of Information and Communication Engineering |
Files in This Item:
File | Description | Size | Format | |
---|---|---|---|---|
01_title.pdf | Attached File | 44.95 kB | Adobe PDF | View/Open |
02_prelim_pages.pdf | 3.12 MB | Adobe PDF | View/Open | |
03_content.pdf | 92.55 kB | Adobe PDF | View/Open | |
04_abstract.pdf | 85.28 kB | Adobe PDF | View/Open | |
05_chapter 1.pdf | 501.23 kB | Adobe PDF | View/Open | |
06_chapter 2.pdf | 371.28 kB | Adobe PDF | View/Open | |
07_chapter 3.pdf | 773.66 kB | Adobe PDF | View/Open | |
08_chapter 4.pdf | 478.37 kB | Adobe PDF | View/Open | |
09_chapter 5.pdf | 2 MB | Adobe PDF | View/Open | |
10_annexures.pdf | 117.07 kB | Adobe PDF | View/Open | |
80_recommendation.pdf | 86.31 kB | Adobe PDF | View/Open |
Items in Shodhganga are licensed under Creative Commons Licence Attribution-NonCommercial-ShareAlike 4.0 International (CC BY-NC-SA 4.0).
Altmetric Badge: