Survey of Mitosis Detection Techniques in Breast Cancer

Abstract

Breast Cancer is one of the most occurring cancer in women's among different types of cancer. Detection of mitosis is a challenging work in breast histopathology images and in mammography. Due to different stages of mitosis, non uniform stain illumination, size and shape variation of cells detection of mitosis is difficult. Different techniques and algorithms are used for detection of mitosis and non mitosis cells. Others additional techniques like fuzzy logic, neuro-fuzzy system, Artificial Neural Network, Convlutional Neural Network and classifiers also used for classification of malignant and non malignant cells based on learned features. Image recognition system using convolution neural network (CNN) used to recognize visual imagery. To detect lesions (damaged tissue) Basian Neural network is also deployed. Deep learning technique also used on mammogram by extracting features from subdivided abnormal classes to normal classes. This paper gives survey and comparative analysis of different techniques used for detection of breast cancer.