A CNN Based Breast Tumor Classifier Using Mendeley BUS Dataset
S.Sri Durga Kameswari1, S.Sri Durga Kameswari2, V Vijayakumar3
1S. Sri Durga Kameswari, Department of ECE, Sathyabama Institute of Science and Technology, Chennai (Tamil Nadu), India.
2S. Sri Durga Kameswari, Department of ECE, GMR Institute of Technology, Rajam (Andhra Pradesh), India.
3Dr. V Vijayakumar, Department of ECE, Sathyabama Institute of Science and Technology, Chennai (Tamil Nadu), India.
Manuscript received on 07 April 2019 | Revised Manuscript received on 20 April 2019 | Manuscript published on 30 April 2019 | PP: 1340-1343 | Volume-8 Issue-6, April 2019 | Retrieval Number: F3878048619/19©BEIESP
Open Access | Ethics and Policies | Cite | Mendeley | Indexing and Abstracting
© The Authors. Blue Eyes Intelligence Engineering and Sciences Publication (BEIESP). This is an open access article under the CC-BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/)
Abstract: Women account to nearly fifty percent of the world population and most of them were being diagnosed with breast cancer. In the last decade, the worldwide breast cancer rate has been increased by more than twenty percent. A tumor may be cancerous or non-cancerous. There is a severe need to categorize the breast tumor for further treatment. Computer aided identification techniques can minimize the number of unnecessary biopsies. In this paper, a deep learning algorithm is proposed to label the tumors as benign or malignant. It uses three layers within it. We have implemented our model on breast ultrasound images. Mendeley dataset was used for this experiment which contains 250 ultrasound images of breast tumor out of which 100 were benign and 150 were of malignant class. The Proposed CNN was trained from scratch, trained with 50 epochs yields test accuracy of approximately 98%.
Keyword: Breast Cancer, Deep Learning, Ultrasound, CNN.
Scope of the Article: Agent-Based Software Engineering