Skip to main navigation Skip to search Skip to main content

Improving Medical Image Segmentation Using Gaze Guidance from Non-Professionals

  • Yinfeng Fang
  • , Wenlong Ma
  • , Xixia Yu
  • , Qingsong Liu
  • , Yuxi Wang
  • , Yong Peng
  • , Dalin Zhou
  • , Zhaojie Ju*
  • *Corresponding author for this work

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

    7 Downloads (Pure)

    Abstract

    Medical image segmentation using artificial intelligence (AI) has significantly improved diagnosis and treatment, enhancing patient outcomes. However, AI's reliance on large amounts of labeled data and expert annotations limits its scalability. This paper explores the use of eye movement data from non-professionals to improve segmentation accuracy. A novel gaze-enhanced image segmentation model (GISM) is proposed, consisting of two main components: a transformer-based gaze feature extraction block (TGFB) and a multi-head gazeattention block (MHGB). The MHGB fuses gaze and image data by using gaze features as queries within a transformer framework. The colon polyp dataset, is applied to assess the model's performance. The outcomes reveal that 1) gaze data from non-professionals enhances segmentation accuracy, 2) TGFB outperforms other methods in fusing gaze information, and 3) TGFB is highly adaptable to encoder/decoder-based segmentation frameworks.

    Original languageEnglish
    Title of host publication2025 10th IEEE International Conference on Advanced Robotics and Mechatronics, ICARM 2025
    PublisherInstitute of Electrical and Electronics Engineers Inc.
    Pages100-105
    Number of pages6
    ISBN (Electronic)9798331503079
    ISBN (Print)9798331503086
    DOIs
    Publication statusPublished - 22 Dec 2025
    Event2025 10th IEEE International Conference on Advanced Robotics and Mechatronics, ICARM 2025 - Portsmouth, United Kingdom
    Duration: 1 Aug 20253 Aug 2025

    Publication series

    NameIEEE ICARM Proceedings
    ISSN (Print)2993-4982
    ISSN (Electronic)2993-4990

    Conference

    Conference2025 10th IEEE International Conference on Advanced Robotics and Mechatronics, ICARM 2025
    Country/TerritoryUnited Kingdom
    CityPortsmouth
    Period1/08/253/08/25

    Fingerprint

    Dive into the research topics of 'Improving Medical Image Segmentation Using Gaze Guidance from Non-Professionals'. Together they form a unique fingerprint.

    Cite this