Modified Hybrid Task Cascade for Lung Nodules Segmentation in CT Images with Guided Anchoring

Shihuai Xu, Huijuan Lu, Minchao Ye, Ke Yan, Wenjie Zhu, Qun Jin

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

As lung cancer continues to threaten human health, Computer-Aided Diagnostic (CAD) plays an increasingly significant role in lung cancer diagnosis, and convolutional neural networks (CNNs) have shown the outstanding performance in image segmentation. In this work, Hybrid Task Cascade (HTC) is used to segment lung nodules that are difficult to find in CT images. Considering that lung nodules are usually quite small, this study integrates Feature Pyramid Network (FPN) into ResNet-50 to make full use of multi-scale feature and improve the segmentation accuracy of small target nodules. In addition, given that existing defects in Region Proposal Network (RPN), which refers to most of generated anchors are irrelevant to target objects, and the conventional method are unaware of the shapes of target objects, this work proposes to use Guided Anchoring to replace RPN in HTC and generate anchors more effectively. Experimental results on the LIDC-IDRI dataset demonstrate that the modified HTC improves the segmentation accuracy of lung nodules.

Original languageEnglish
Title of host publicationProceedings of the 2020 12th International Conference on Machine Learning and Computing, ICMLC 2020
PublisherAssociation for Computing Machinery
Pages433-438
Number of pages6
ISBN (Electronic)9781450376426
DOIs
Publication statusPublished - 2020 Feb 15
Event12th International Conference on Machine Learning and Computing, ICMLC 2020 - Shenzhen, China
Duration: 2020 Feb 152020 Feb 17

Publication series

NameACM International Conference Proceeding Series

Conference

Conference12th International Conference on Machine Learning and Computing, ICMLC 2020
CountryChina
CityShenzhen
Period20/2/1520/2/17

Keywords

  • CT images
  • Guided Anchoring
  • Hybrid Task Cascade
  • lung nodules segmentation

ASJC Scopus subject areas

  • Software
  • Human-Computer Interaction
  • Computer Vision and Pattern Recognition
  • Computer Networks and Communications

Fingerprint Dive into the research topics of 'Modified Hybrid Task Cascade for Lung Nodules Segmentation in CT Images with Guided Anchoring'. Together they form a unique fingerprint.

Cite this