University of Leicester
Browse
- No file added yet -

Light Annotation Fine Segmentation: Histology Image Segmentation based on VGG Fusion with Global Normalisation CAM

Download (2.3 MB)
conference contribution
posted on 2022-08-12, 09:25 authored by Y Li, Y Wang, L Dong, J Ye, L Wang, R Ge, Huiyu Zhou, Q Zhang

Deep learning has been widely used to segment tumour regions in stained histopathology images. However, precise annotations are expensive and labour-consuming. To reduce the manual annotation workload, we propose a light annotation-based fine-level segmentation approach for histology images based on a VGG-based Fusion network with Global Normalisation CAM. The experts are only required to provide a rough segmentation annotation on the images, and then accurate fine-level segmentation boundaries can be produced using this method. To validate the proposed approach, three datasets with rough and fine quality segmentation annotation are built. The fine quality labels are used only as ground truth in evaluation. The VFGN-CAM method includes three main components: an annotation enhancement to boost boundary accuracy and model generalisability; a VGG Fusion module that integrates multi-scale tumour features; and a Global Normalisation CAM module that combines local and global gradient information of tumour regions. Our VGG fusion and Global Normalisation CAM outperform the existing methods with a Dice of 84.188%. The final improvement for our proposed methods against the original rough labels is around 22.8%.

History

Author affiliation

School of Computing and Mathematical Sciences, University of Leicester

Source

The 1st Workshop on Computational Mathematics Modeling in Cancer Analysis, A MICCAI 2022 Workshop, September 18th, 2022, Singapore

Version

  • AM (Accepted Manuscript)

Published in

Computational Mathematics Modeling in Cancer Analysis. CMMCA 2022.

Pagination

121-130

Publisher

Springer

isbn

978-3-031-17266-3

Acceptance date

2022-08-05

Copyright date

2022

Available date

2022-09-19

Book series

Lecture Notes in Computer Science, vol 13574

Temporal coverage: start date

2022-09-18

Temporal coverage: end date

2022-09-18

Language

en

Usage metrics

    University of Leicester Publications

    Categories

    No categories selected

    Exports

    RefWorks
    BibTeX
    Ref. manager
    Endnote
    DataCite
    NLM
    DC