EUVIMEDEUVIMED entdecken
Crossref

Weakly supervised semantic segmentation of histological tissue via attention accumulation and pixel-level contrast learning

Yongqi Han, Lianglun Cheng, Guoheng Huang iD, Guo Zhong, Jiahua Li, Xiaochen Yuan, Hongrui Liu, Jiao Li, Jian Zhou, Muyan Cai

Physics in Medicine & Biology · 2023

DOI: 10.1088/1361-6560/acaeee ↗

Kurz gefasst

Fachartikel in der Zeitschrift Physics in Medicine & Biology, 2023. Verfasst von Yongqi Han und 9 weiteren Autor:innen. Beteiligte Einrichtung: School of Computer Science and Technology, Guangdong University of Technology, Guangzhou 510006, People's Republic of China. Hauptthemen laut MEDLINE: Adenocarcinoma of Lung, Lung Neoplasms. Die Arbeit wurde bisher 2-mal zitiert (Crossref) und nennt 78 Literaturangaben. Gefördert von: Guangdong Provincial Key Laboratory of Cyber-Physical System, Guangdong R&D plan projects in key areas, Key-Area Research and Development Program of Guangdong Province.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Objective . Histopathology image segmentation can assist medical professionals in identifying and diagnosing diseased tissue more efficiently. Although fully supervised segmentation models have excellent performance, the annotation cost is extremely expensive. Weakly supervised models are widely used in medical image segmentation due to their low annotation cost. Nevertheless, these weakly supervised models have difficulty in accurately locating the boundaries between different classes of regions in pathological images, resulting in a high rate of false alarms Our objective is to design a weakly supervised segmentation model to resolve the above problems. Approach . The segmentation model is divided into two main stages, the generation of pseudo labels based on class residual attention accumulation network (CRAANet) and the semantic segmentation based on pixel feature space construction network (PFSCNet). CRAANet provides attention scores for each class through the class residual attention module, while the Attention Accumulation (AA) module overlays the attention feature maps generated in each training epoch. PFSCNet employs a network model containing an inflated convolutional residual neural network and a multi-scale feature-aware module as the segmentation backbone, and proposes dense energy loss and pixel clustering modules are based on contrast learning to solve the pseudo-labeling-inaccuracy problem. Main results . We validate our method using the lung adenocarcinoma (LUAD-HistoSeg) dataset and the breast cancer (BCSS) dataset. The results of the experiments show that our proposed method outperforms other state-of-the-art methods on both datasets in several metrics. This suggests that it is capable of performing well in a wide variety of histopathological image segmentation tasks. Significance . We propose a weakly supervised semantic segmentation network that achieves approximate fully supervised segmentation performance even in the case of incomplete labels. The proposed AA and pixel-level contrast learning also make the edges more accurate and can well assist pathologists in their research.

Schlagwörter

MeSH-Schlagwörter (MEDLINE)

Adenocarcinoma of LungLung NeoplasmsHumansSemanticsBenchmarkingCluster AnalysisImage Processing, Computer-Assisted

Stichwörter der Autor:innen

attention accumulationcontrast learninghistological tissueweakly supervised semantic segmentation

Autor:innen und Institutionen

  • School of Computer Science and Technology, Guangdong University of Technology, Guangzhou 510006, People's Republic of China.
  • School of Information Science and Technology, Guangdong University of Foreign Studies, Guangzhou 510420, People's Republic of China.
  • Faculty of Applied Sciences, Macao Polytechnic University, Macao 999078, People's Republic of China.
  • Department of Industrial and Systems Engineering, San Jose State University, CA 95192, United States of America.
  • Department of Radiology, Sun Yat-sen University Cancer Center; State Key Laboratory of Oncology in South China, Collaborative Innovation Center for Cancer Medicine, Guangzhou 510060, People's Republic of China.
  • Department of Medical Imaging, Sun Yat-sen University Cancer Center; State Key Laboratory of Oncology in South China; Collaborative Innovation Center for Cancer Medicine, Guangzhou 510060, People's Republic of China.

ORCID-Profile: Guoheng Huang

Förderung

  • Guangdong Provincial Key Laboratory of Cyber-Physical System
  • Guangdong R&D plan projects in key areas
  • Key-Area Research and Development Program of Guangdong Province
  • Science and technology research in key areas in Foshan
  • National Statistical Science Research Project of China
  • Guangzhou Science and Technology Project

Zitieren

APA

Han, Y., Cheng, L., Huang, G., Zhong, G., Li, J., Yuan, X., Liu, H., Li, J., Zhou, J., & Cai, M. (2023). Weakly supervised semantic segmentation of histological tissue via attention accumulation and pixel-level contrast learning. Physics in Medicine & Biology. https://doi.org/10.1088/1361-6560/acaeee

Vancouver

Han Y, Cheng L, Huang G, Zhong G, Li J, Yuan X, et al. Weakly supervised semantic segmentation of histological tissue via attention accumulation and pixel-level contrast learning. Physics in Medicine & Biology. 2023. doi:10.1088/1361-6560/acaeee

BibTeX
@article{Han2023,
  author = {Han, Yongqi and Cheng, Lianglun and Huang, Guoheng and Zhong, Guo and Li, Jiahua and Yuan, Xiaochen and Liu, Hongrui and Li, Jiao and Zhou, Jian and Cai, Muyan},
  title = {Weakly supervised semantic segmentation of histological tissue via attention accumulation and pixel-level contrast learning},
  journal = {Physics in Medicine & Biology},
  year = {2023},
  doi = {10.1088/1361-6560/acaeee}
}
Kostenlos starten →