EUVIMEDEUVIMED entdecken
Crossref

A multimodal image-based automated method for macular hole boundary detection

Qingxin Yuan, Rong Tan, Jianxin Shen, Fen Zhou, Jianguo Xu

Artificial Intelligence in Health · 23.07.2026 · 026220053

DOI: 10.36922/aih026220053 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Artificial Intelligence in Health, online erschienen am 23.07.2026. Verfasst von Qingxin Yuan und 4 weiteren Autor:innen.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Accurate detection of the macular hole (MH) boundary is crucial for diagnosis and treatment planning. Manual annotation by ophthalmologists is time-consuming, poorly reproducible, and subject to substantial inter-observer variability. To address these challenges, this paper proposes an automated MH boundary detection method based on multimodal images. The method mainly consists of the following stages: (i) automatically localizing the scanning line center in the fundus infrared (IR) image using image processing techniques to provide a reference point for subsequent cross-modal mapping; (ii) introducing the YOLOv8-pose keypoint detection model into the task of MH highest point detection, achieving its automated localization in fundus optical coherence tomography (OCT) B-scan images; (iii) constructing a cross-modal mapping algorithm that, through a combination of position transformation and rotation, facilitates the mapping of the MH highest point from OCT B-scan images to IR images; (iv) designing a boundary fitting scheme that connects all mapped points with reference to adjacent scanning lines to form a closed polygon, thereby achieving MH boundary detection. Finally, a series of quantitative and qualitative experiments were conducted to evaluate the performance of the YOLOv8-pose-based keypoint detection model on public and clinical MH datasets for highest-point detection and to validate the effectiveness of the proposed method on a clinical dataset for MH boundary detection. This study presents a new approach to automated MH boundary detection and may facilitate the accurate diagnosis and treatment planning of MH.

Zitieren

APA

Yuan, Q., Tan, R., Shen, J., Zhou, F., & Xu, J. (2026). A multimodal image-based automated method for macular hole boundary detection. Artificial Intelligence in Health, 0(0), 026220053. https://doi.org/10.36922/aih026220053

Vancouver

Yuan Q, Tan R, Shen J, Zhou F, Xu J. A multimodal image-based automated method for macular hole boundary detection. Artificial Intelligence in Health. 2026;0(0):026220053. doi:10.36922/aih026220053

BibTeX
@article{Yuan2026,
  author = {Yuan, Qingxin and Tan, Rong and Shen, Jianxin and Zhou, Fen and Xu, Jianguo},
  title = {A multimodal image-based automated method for macular hole boundary detection},
  journal = {Artificial Intelligence in Health},
  year = {2026},
  volume = {0},
  number = {0},
  pages = {026220053},
  doi = {10.36922/aih026220053}
}
Kostenlos starten →