We gratefully acknowledge support from
the Simons Foundation and member institutions.
Full-text links:

Download:

Current browse context:

eess.IV

Change to browse by:

References & Citations

Bookmark

(what is this?)
CiteULike logo BibSonomy logo Mendeley logo del.icio.us logo Digg logo Reddit logo ScienceWISE logo

Electrical Engineering and Systems Science > Image and Video Processing

Title: Deep Multi-Scale U-Net Architecture and Label-Noise Robust Training Strategies for Histopathological Image Segmentation

Abstract: Although the U-Net architecture has been extensively used for segmentation of medical images, we address two of its shortcomings in this work. Firstly, the accuracy of vanilla U-Net degrades when the target regions for segmentation exhibit significant variations in shape and size. Even though the U-Net already possesses some capability to analyze features at various scales, we propose to explicitly add multi-scale feature maps in each convolutional module of the U-Net encoder to improve segmentation of histology images. Secondly, the accuracy of a U-Net model also suffers when the annotations for supervised learning are noisy or incomplete. This can happen due to the inherent difficulty for a human expert to identify and delineate all instances of specific pathology very precisely and accurately. We address this challenge by introducing auxiliary confidence maps that emphasize less on the boundaries of the given target regions. Further, we utilize the bootstrapping properties of the deep network to address the missing annotation problem intelligently. In our experiments on a private dataset of breast cancer lymph nodes, where the primary task was to segment germinal centres and sinus histiocytosis, we observed substantial improvement over a U-Net baseline based on the two proposed augmentations.
Comments: 12 pages, 4 figures , 2 tables ,Added Attention UNet Results, Added Sinus and Germinal Center overlay images, Modified paper format, Fixed Title typos
Subjects: Image and Video Processing (eess.IV); Computer Vision and Pattern Recognition (cs.CV)
Cite as: arXiv:2205.01777 [eess.IV]
  (or arXiv:2205.01777v2 [eess.IV] for this version)

Submission history

From: NIkhil Cherian Kurian [view email]
[v1] Tue, 3 May 2022 21:00:44 GMT (4055kb,D)
[v2] Sat, 13 Aug 2022 21:31:08 GMT (13633kb,D)

Link back to: arXiv, form interface, contact.