Breast cancer detection Using Convolutional Neural Networks for mammogram imaging system

Y. J. Tan, K. S. Sim, F. F. Ting

Research output: Chapter in Book/Report/Conference proceedingConference PaperResearchpeer-review

98 Citations (Scopus)


In this paper, breast cancer detection using convolutional neural network for mammogram imaging system is proposed to classify mammogram image into normal, benign(non-cancerous abnormality) and malignant (cancerous abnormality). Breast Cancer detection Using Convolutional Neural Networks (BCDCNN) is aimed to speed up the diagnosis process by assisting specialist to diagnosis and classification the breast cancer. A series of mammogram images are used to carry out preprocessing to convert a human visual image into a computer visual image and adjust suitable parameter for the CNN classifier. After that, all changed images are assigned into CNN classifier as training source. The CNN classifier will then produce a model to recognize the mammogram image. By comparing BCDCNN method with Mammogram Classification Using Convolutional Neural Networks (MCCNN), BCDCNN has improved the accuracy toward classification on the mammogram images. Thus, the results show that the proposed method has higher accuracy than other existing methods, mass only and all argument have been increased from 0.75 to 0.8585 and 0.608974 to 0.8271 accuracy.

Original languageEnglish
Title of host publicationProceedings of 2017 International Conference on Robotics, Automation and Sciences ICORAS 2017
EditorsSim Kwok Swee
Place of PublicationPiscataway NJ USA
PublisherIEEE, Institute of Electrical and Electronics Engineers
Number of pages5
ISBN (Electronic)9781538619087
ISBN (Print)9781538619094
Publication statusPublished - 2017
Externally publishedYes
EventInternational Conference on Robotics, Automation and Sciences 2017 - Melaka, Malaysia
Duration: 27 Nov 201729 Nov 2017 (Proceedings)


ConferenceInternational Conference on Robotics, Automation and Sciences 2017
Abbreviated titleICORAS 2017
Internet address

Cite this