Deep learning for damaged tissue detection and segmentation in Ki-67 brain tumor specimens based on the U-net model

  title={Deep learning for damaged tissue detection and segmentation in Ki-67 brain tumor specimens based on the U-net model},
  author={Zaneta Swiderska-Chadaj and Tomasz Markiewicz and Jaime Gallego and Gloria Bueno and Bartlomiej Grala and Malgorzata Lorent},
  journal={Bulletin of The Polish Academy of Sciences-technical Sciences},
Bull. Pol. Ac.: Tech. 66(6) 2018 Abstract. The pathologists follow a systematic and partially manual process to obtain histological tissue sections from the biological tissue extracted from patients. This process is far from being perfect and can introduce some errors in the quality of the tissue sections (distortions, deformations, folds and tissue breaks). In this paper, we propose a deep learning (DL) method for the detection and segmentation of these damaged regions in whole slide images… 

Figures and Tables from this paper

Generating Region of Interests for Invasive Breast Cancer in Histopathological Whole-Slide-Image
A CNN based architecture for generating ROIs through segmentation is proposed based on transfer learning and the use of dilated convolutions for the detection of invasive breast cancer regions.
U-Net based frames partitioning and volumetric analysis for kidney detection in tomographic images
A method of image partitioning into frames, thereby providing only a fragment of the image to the network input, and the volumetric coherent area analysis technique is presented, which facilitates removing errors in the form of redundant areas generated by the neural network.
A State-of-the-art Survey of U-Net in Microscopic Image Analysis: from Simple Usage to Structure Mortification
The improved methods of U-Net are summarized and the existing significances of image segmentation techniques and their improvements that has introduced over the years are listed.
Segmentation of bone structures with the use of deep learning techniques
The authors trained and compared four different models of deep neural networks to perform the segmentation task of three following classes: background, patient outline and bones to obtain the most exact segmentation results.
Deep Learning for Histopathological Image Analysis
All current methods applied to histopathological image analysis will be presented as well as the future technological issues and challenges of this discipline.
piNET: An Automated Proliferation Index Calculator Framework for Ki67 Breast Cancer Images
In this work, a novel proliferation index (PI) calculator for Ki67 images called piNET is proposed. It is successfully tested on four datasets, from three scanners comprised of patches, tissue
Automatic Segmentation of Immunohistochemical Images Based on U-net Architecture
The program implemented an application for automatic segmentation of immunohistochemical images, which allowed to improve the quality of segmentation based on the training sample and presented an image processing algorithm, which includes the stage of pre-processing and automatic segmentations.


A Deep Learning Pipeline to Delineate Proliferative Areas of Intracranial Tumors in Digital Slides
A deep learning based pipeline to delineate areas of tumor in meningioma and oligodendroglioma specimens stained with Ki-67 marker is presented and has the potential to objectively pre-process slides for proliferative index quantification.
Semantic segmentation of microscopic images of H&E stained prostatic tissue using CNN
An algorithm for semantic segmentation of the microscopic images of H&E stained prostate tissue into Background, Stroma, Epithelial Cytoplasm and Nuclei is developed, based on deep learning, or more specifically a convolutional neural network.
Multi-loss convolutional networks for gland analysis in microscopy
A novel multi-objective learning method is proposed that optimizes a single unified deep fully convolutional neural network with two distinct loss functions with an improvement of 6% over classification-only models.
Automatic Brain Tumor Detection and Segmentation Using U-Net Based Fully Convolutional Networks
This study proposes a fully automatic method for brain tumor segmentation, which is developed using U-Net based deep convolutional networks, which was evaluated on Multimodal Brain Tumor Image Segmentation (BRATS 2015) datasets, showing that it can obtain promising segmentation efficiently.
Nuclei segmentation in histopathology images using deep neural networks
This work presents a fully automated workflow to segment nuclei from histopathology image data by using deep neural networks trained from a set of manually annotated images and by processing the posterior probability maps in order to split jointly segmented nuclei.
Fully convolutional networks with double-label for esophageal cancer image segmentation by self-transfer learning
This paper focuses on the semantic segmentation task of microvascular morphological types on narrowband images to aid clinical examination of esophageal cancer and shows that the proposed approach is able to assist clinical diagnosis to a certain extent.
U-Net: Convolutional Networks for Biomedical Image Segmentation
It is shown that such a network can be trained end-to-end from very few images and outperforms the prior best method (a sliding-window convolutional network) on the ISBI challenge for segmentation of neuronal structures in electron microscopic stacks.
Eliminating tissue-fold artifacts in histopathological whole-slide images for improved image-based prediction of cancer grade
After eliminating tissue folds, the performance of cancer-grade prediction models improved by 5% and 1% in OvCa and KiCa, respectively, and the proposed connectivity-based method is more effective in detecting tissue folds compared to other methods.
Glomerulus Classification and Detection Based on Convolutional Neural Networks
The results indicate that this technique is suitable for correct Glomerulus detection in Whole Slide Images (WSI), showing robustness while reducing false positive and false negative detections.
Image analysis and machine learning in digital pathology: Challenges and opportunities