Deep Learning Layer Convolutional Neural Network (CNN) Scheme for Cancer Image

Joint Conference on Green Engineering Technology & Applied Computing 2019

工业技术（总论）;计算机科学

Zainudin, Z.^1 ; Shamsuddin, S.M.^1 ; Hasan, S.^1

School of Computing, Faculty of Engineering, Universiti Teknologi Malaysia, Malaysia^1

关键词: Breast Cancer;    Cancer cells;    Cancer classification;    Cancer detection;    Convolutional neural network;    Digital image;    Female breast;    Small objects;   

Others  :  https://iopscience.iop.org/article/10.1088/1757-899X/551/1/012039/pdf
DOI  :  10.1088/1757-899X/551/1/012039

来源: IOP

【 摘 要 】

Recent years, in medical image especially cancer detection used whole slide digital scanners, called as histopathology image (images of tissues and cell) where it can now be keep in digital images. Consequently, using Deep Learning will help pathologist in cancer detection (cancer cell known as mitosis). In this paper, we are using Deep Learning Layer Convolutional Neural Network (CNN) for cancer classification using histopathology image and used AMIDA dataset which are related on female breast cancer dataset. Mitosis is an important parameter for the prognosis/diagnosis of breast cancer. However, using histopathology image for cancer detection is a challenging problem that needs a deeper investigations. This problem occurs when to classify mitosis because mitosis is small objects with a large variety of shapes, and they can thus be easily confused with some other objects or artefacts present in the image. In this paper, the objective to find the suitable layer for Deep Learning Convolutional Neural Network and reduce the loss rate.