ScaleNet: Rethinking Feature Interaction from a Scale-Wise Perspective for Medical Image Segmentation

  • Yu Feng
  • , Tai Ma
  • , Hao Zeng
  • , Zhengke Xu
  • , Suwei Zhang
  • , Ying Wen*
  • *Corresponding author for this work

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

Abstract

Recently, vision transformers have become outstanding segmentation structures for their remarkable global modeling capability. In current transformer-based models for medical image segmentation, convolutional layers are often replaced by transformers, or transformers are added to the deepest layer of the encoder to learn the global context. However, for the extracted multi-scale feature information, most existing methods tend to ignore the multi-scale dependencies, which leads to inadequate feature learning and fails to produce rich feature representations. In this paper, we propose ScaleNet from the perspective of feature interaction at different scales that can alleviate mentioned problems. Specifically, our approach consists of two multi-scale feature interaction modules: the spatial scale interaction (SSI) and the channel scale interaction (CSI). SSI uses a transformer to aggregate patches from different scale features to enhance the feature representations at the spatial scale. CSI uses a 1D convolutional layer and a fully connected layer to perform a global fusion of multi-level features at the channel scale. The combination of CSI and SSI enables ScaleNet to emphasize multi-scale dependencies and effectively resolve complex scale variations.

Original languageEnglish
Title of host publicationAdvances in Computer Graphics - 40th Computer Graphics International Conference, CGI 2023, Proceedings
EditorsBin Sheng, Lei Bi, Jinman Kim, Nadia Magnenat-Thalmann, Daniel Thalmann
PublisherSpringer Science and Business Media Deutschland GmbH
Pages222-236
Number of pages15
ISBN (Print)9783031500770
DOIs
StatePublished - 2024
Event40th Computer Graphics International Conference, CGI 2023 - Shanghai, China
Duration: 28 Aug 20231 Sep 2023

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume14498 LNCS
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Conference

Conference40th Computer Graphics International Conference, CGI 2023
Country/TerritoryChina
CityShanghai
Period28/08/231/09/23

Keywords

  • medical image segmentation
  • multi-organ and skin lesion segmentation tasks
  • multi-scale feature interaction
  • transformer-based method

Fingerprint

Dive into the research topics of 'ScaleNet: Rethinking Feature Interaction from a Scale-Wise Perspective for Medical Image Segmentation'. Together they form a unique fingerprint.

Cite this