Caln: Channel Attention LSTM Network for Pituitary Segmentation in Dynamic Contrast-Enhanced MRI

Jingyi Liu, Zhaoze Sun, Qing Guo, Bing Liu, Tianyu Fu*, Guolin Ma, Hong Song*, Jian Yang

*Corresponding author for this work

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

Abstract

Automatic pituitary segmentation in MR images is crucial for pituitary-related diseases diagnosis, yet it is challenging due to limited spatial and temporal resolutions of the image and the small size of the gland. In this study, we introduce the Channel Attention Long Short-Term Memory Network (CALN), integrating Long Short-Term Memory (LSTM) units into the bottleneck of a symmetrical encoder-decoder structure. We employ channel-wise attention to fuse features extracted from different time phases based on their semantic importance. Notably, CALN achieves remarkable performance with only two downsampling layers. To evaluate the effectiveness of CALN, we conducted extensive experiments and compared it with state-of-the-art methods for pituitary segmentation. Our results show that CALN outperforms UNETR by 3.2% and SwinUNETR by 1.2% in terms of the dice score. The findings confirm the improved accuracy and computational efficiency of our method for pituitary segmentation.

Original languageEnglish
Title of host publicationIEEE International Symposium on Biomedical Imaging, ISBI 2024 - Conference Proceedings
PublisherIEEE Computer Society
ISBN (Electronic)9798350313338
DOIs
Publication statusPublished - 2024
Event21st IEEE International Symposium on Biomedical Imaging, ISBI 2024 - Athens, Greece
Duration: 27 May 202430 May 2024

Publication series

NameProceedings - International Symposium on Biomedical Imaging
ISSN (Print)1945-7928
ISSN (Electronic)1945-8452

Conference

Conference21st IEEE International Symposium on Biomedical Imaging, ISBI 2024
Country/TerritoryGreece
CityAthens
Period27/05/2430/05/24

Keywords

  • channel-wise attention
  • LSTM
  • medical image segmentation
  • pituitary
  • U-Net

Fingerprint

Dive into the research topics of 'Caln: Channel Attention LSTM Network for Pituitary Segmentation in Dynamic Contrast-Enhanced MRI'. Together they form a unique fingerprint.

Cite this