Breast Cancer Histological Images Nuclei Segmentation using Mask Regional Convolutional Neural Network
- 27 September 2020
- conference paper
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 1-6
Abstract
the breast cancer cases have increased globally, in the last two decades. Analysis of histological images acquired through biopsy of the breast tissues is thought to be the most reliable way to assess if any cells show signs of malignancy. With the advent of deep neural networks, we are now able to diagnose the disease with near perfect accuracy instead of conventional techniques of mammography for breast cancer detection; however, work done in this area considers nuclei segmentation. This work is focussed on extending the applicability of deep learning technology where network is trained to perform segmentation and classify data as one of the four classes: normal, benign, in-situ and invasive. The classification accuracy achieved is 98.16% and the segmentation accuracy calculated using overlap coefficient from mean intersection over union is 86.39% on the ICIAR2018 Breast Cancer dataset.Keywords
Funding Information
- Ministry of Education
This publication has 17 references indexed in Scilit:
- Diagnostic Assessment of Deep Learning Algorithms for Detection of Lymph Node Metastases in Women With Breast CancerJAMA, 2017
- Accurate and reproducible invasive breast cancer detection in whole-slide images: A Deep Learning approach for quantifying tumor extentScientific Reports, 2017
- Signal Processing Techniques and Computer-Aided Detection Systems for Diagnosis of Breast Cancer – A Review PaperIndian Journal of Science and Technology, 2017
- Breast cancer histopathological image classification using Convolutional Neural Networks2016
- Fast R-CNN2015
- Region segmentation in histopathological breast cancer images using deep convolutional neural network2015
- Deep convolutional activation features for large scale Brain Tumor histopathology image classification and segmentation2015
- Breast Cancer Histopathology Image Analysis: A ReviewIEEE Transactions on Biomedical Engineering, 2014
- Multi-Dimensional Tumor Detection in Automated Whole Breast Ultrasound Using Topographic WatershedIEEE Transactions on Medical Imaging, 2014
- Incorporating domain knowledge for tubule detection in breast histopathology using O'Callaghan neighborhoods2011