Multi-rater Prism:Learning self-calibrated medical image segmentation from multiple raters

Junde Wu ¹Huihui Fang ²Jiayuan Zhu ³Yu Zhang ⁴Xiang Li ⁵Yuanpei Liu ⁶Huiying Liu ⁷Yueming Jin ⁸Weimin Huang ⁷Qi Liu ⁹Cen Chen ⁹Yanfei Liu ¹⁰Lixin Duan ¹¹Yanwu Xu ¹²Li Xiao ¹³Weihua Yang ¹⁴Yue Liu¹⁰

扫码查看

作者信息

1. School of Future Technology,South China University of Technology,Guangzhou 511442,China;Pazhou Lab,Guangzhou 510320,China;The University of Oxford,Oxford OX14AL,UK
2. School of Future Technology,South China University of Technology,Guangzhou 511442,China;Pazhou Lab,Guangzhou 510320,China;Cardiovascular Disease Center,Xiyuan Hospital of China Academy of Chinese Medical Sciences,Beijing 100091,China
3. The University of Oxford,Oxford OX14AL,UK
4. State Key Laboratory of Pulsed Power Laser Technology,College of Electronic Engineering,National University of Defense Technology,Hefei 230037,China
5. Shenzhen Institute for Advanced Study,University of Electronic Science and Technology of China,Shenzhen 518110,China
6. The University of Hong Kong,Hong Kong 999077,China
7. Institute for Infocomm Research,A*STAR,Singapore 138632,Singapore
8. National University of Singapore,Singapore 119276,Singapore
9. School of Future Technology,South China University of Technology,Guangzhou 511442,China
10. Cardiovascular Disease Center,Xiyuan Hospital of China Academy of Chinese Medical Sciences,Beijing 100091,China
11. Shenzhen Institute for Advanced Study,University of Electronic Science and Technology of China,Shenzhen 518110,China;Sichuan Provincial People's Hospital,University of Electronic Science and Technology of China,Chengdu 611731,China
12. School of Future Technology,South China University of Technology,Guangzhou 511442,China;Pazhou Lab,Guangzhou 510320,China
13. Sichuan Provincial People's Hospital,University of Electronic Science and Technology of China,Chengdu 611731,China
14. Shenzhen Eye Hospital, Jinan University, Shenzhen 518040, China
折叠

Abstract

In medical image segmentation,it is often necessary to collect opinions from multiple experts to make the final decision.This clinical routine helps to mitigate individual bias.However,when data is annotated by multiple experts,standard deep learning models are often not applicable.In this paper,we propose a novel neural network framework called Multi-rater Prism(MrPrism)to learn medical image segmenta-tion from multiple labels.Inspired by iterative half-quadratic optimization,MrPrism combines the task of assigning multi-rater confidences and calibrated segmentation in a recurrent manner.During this pro-cess,MrPrism learns inter-observer variability while taking into account the image's semantic properties and finally converges to a self-calibrated segmentation result reflecting inter-observer agreement.Specifically,we propose Converging Prism(ConP)and Diverging Prism(DivP)to iteratively process the two tasks.ConP learns calibrated segmentation based on multi-rater confidence maps estimated by DivP,and DivP generates multi-rater confidence maps based on segmentation masks estimated by ConP.Experimental results show that the two tasks can mutually improve each other through this recur-rent process.The final converged segmentation result of MrPrism outperforms state-of-the-art(SOTA)methods for a wide range of medical image segmentation tasks.The code is available at https://github.-com/WuJunde/MrPrism.

Key words

Medical image segmentation/Multiple raters/Self-calibration/Half-quadratic algorithm

引用本文复制引用

基金项目

Excellent Young Science and Technology Talent Cultivation Special Project of China Academy of Chinese Medical Sciences(CI2023D006)

National Natural Science Foundation of China(82121003)

National Natural Science Foundation of China(82022076)

Beijing Natural Science Foundation(2190023)

Shenzhen Fundamental Research Program(JCYJ20220818103207015)

Guangdong Provincial Key Laboratory of Human Digital Twin(2022B1212010004)

出版年

2024

科学通报(英文版)

中国科学院

科学通报(英文版)

CSTPCD

ISSN：1001-6538

段落导航