Tumor segmentation in whole-slide histology images using deep learning

The paper addresses the problem of segmentation of malignant tumors in large whole-slide histology images in the context of computer-assisted diagnosis of breast cancer. The method presented in this study is based on image classification procedure of norm/tumor type. The procedure calculates probabi...

Full description

Saved in:
Bibliographic Details
Main Authors: V. A. Kovalev, V. A. Liauchuk, A. A. Kalinovski, M. V. Fridman
Format: Article
Language:Russian
Published: National Academy of Sciences of Belarus, the United Institute of Informatics Problems 2019-06-01
Series:Informatika
Subjects:
Online Access:https://inf.grid.by/jour/article/view/749
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1849336261544771584
author V. A. Kovalev
V. A. Liauchuk
A. A. Kalinovski
M. V. Fridman
author_facet V. A. Kovalev
V. A. Liauchuk
A. A. Kalinovski
M. V. Fridman
author_sort V. A. Kovalev
collection DOAJ
description The paper addresses the problem of segmentation of malignant tumors in large whole-slide histology images in the context of computer-assisted diagnosis of breast cancer. The method presented in this study is based on image classification procedure of norm/tumor type. The procedure calculates probability of belonging of each particular elementary image region of 256×256 pixels to the "tumor" class, which are isolated by corresponding sliding-window technique. The procedure capitalizes on convolutional neural networks and Deep Learning methods. The neural networks being employed were trained on a representative dataset of 600 000 fragments sampled from whole slide images and representing the morphological and colorimetric variability of two classes. The resultant probability maps were post-processed using conventional image processing algorithms to obtain the final binary masks of pathological regions. The proposed algorithm of segmentation of whole slide histological images can be used in computerized diagnosis of cancer for detection and segmentation of malignant tumors.
format Article
id doaj-art-385b565477ec4192978f825dc6e6b52e
institution Kabale University
issn 1816-0301
language Russian
publishDate 2019-06-01
publisher National Academy of Sciences of Belarus, the United Institute of Informatics Problems
record_format Article
series Informatika
spelling doaj-art-385b565477ec4192978f825dc6e6b52e2025-08-20T03:45:02ZrusNational Academy of Sciences of Belarus, the United Institute of Informatics ProblemsInformatika1816-03012019-06-011621826831Tumor segmentation in whole-slide histology images using deep learningV. A. Kovalev0V. A. Liauchuk1A. A. Kalinovski2M. V. Fridman3The United Institute of Informatics Problems of the National Academy of Sciences of BelarusThe United Institute of Informatics Problems of the National Academy of Sciences of BelarusThe United Institute of Informatics Problems of the National Academy of Sciences of BelarusMinsk City Clinical Oncologic DispensaryThe paper addresses the problem of segmentation of malignant tumors in large whole-slide histology images in the context of computer-assisted diagnosis of breast cancer. The method presented in this study is based on image classification procedure of norm/tumor type. The procedure calculates probability of belonging of each particular elementary image region of 256×256 pixels to the "tumor" class, which are isolated by corresponding sliding-window technique. The procedure capitalizes on convolutional neural networks and Deep Learning methods. The neural networks being employed were trained on a representative dataset of 600 000 fragments sampled from whole slide images and representing the morphological and colorimetric variability of two classes. The resultant probability maps were post-processed using conventional image processing algorithms to obtain the final binary masks of pathological regions. The proposed algorithm of segmentation of whole slide histological images can be used in computerized diagnosis of cancer for detection and segmentation of malignant tumors.https://inf.grid.by/jour/article/view/749whole-slide histology imagesdeep learningconvolutional neural networksimage segmentationcomputerized cancer diagnosis
spellingShingle V. A. Kovalev
V. A. Liauchuk
A. A. Kalinovski
M. V. Fridman
Tumor segmentation in whole-slide histology images using deep learning
Informatika
whole-slide histology images
deep learning
convolutional neural networks
image segmentation
computerized cancer diagnosis
title Tumor segmentation in whole-slide histology images using deep learning
title_full Tumor segmentation in whole-slide histology images using deep learning
title_fullStr Tumor segmentation in whole-slide histology images using deep learning
title_full_unstemmed Tumor segmentation in whole-slide histology images using deep learning
title_short Tumor segmentation in whole-slide histology images using deep learning
title_sort tumor segmentation in whole slide histology images using deep learning
topic whole-slide histology images
deep learning
convolutional neural networks
image segmentation
computerized cancer diagnosis
url https://inf.grid.by/jour/article/view/749
work_keys_str_mv AT vakovalev tumorsegmentationinwholeslidehistologyimagesusingdeeplearning
AT valiauchuk tumorsegmentationinwholeslidehistologyimagesusingdeeplearning
AT aakalinovski tumorsegmentationinwholeslidehistologyimagesusingdeeplearning
AT mvfridman tumorsegmentationinwholeslidehistologyimagesusingdeeplearning