EUVIMEDEUVIMED entdecken
Crossref

Anatomy-Aware Text-Visual Fusion with Dual-Perspective Prompts for Fine-Grained Lumbar Spine Segmentation

Sheng Lian, Jianlong Cai, Dengfeng Pan, Guang-Yong Chen, Hao Xu, Fan Zhang, Guodong Fan, Jialun Pei, Shuo Li

International Journal of Computer Vision · 2026-05 · 134 (5)

DOI: 10.1007/s11263-026-02852-w ↗

Kurz gefasst

Fachartikel in der Zeitschrift International Journal of Computer Vision, online erschienen am 21.04.2026. Verfasst von Sheng Lian und 8 weiteren Autor:innen. Die Arbeit wurde bisher 5-mal zitiert (Crossref) und nennt 57 Literaturangaben.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Accurate lumbar spine segmentation is crucial for diagnosing spinal disorders. Existing methods typically use coarse-grained segmentation strategies that lack the fine details needed for precise diagnosis. Additionally, their reliance on visual-only models hinders the capture of anatomical semantics, leading to misclassified categories and poor segmentation details. To address these limitations, we present ATM-Net, an innovative framework that employs an anatomy-aware, text-guided, multi-modal fusion mechanism for fine-grained segmentation of lumbar substructures, i.e., vertebrae (VBs), intervertebral discs (IDs), and spinal canal (SC). ATM-Net adopts the Anatomy-aware Text Prompt Generator (ATPG) to adaptively convert image annotations into anatomy-aware prompts in different views. These insights are further integrated with image features via the Holistic Anatomy-aware Semantic Fusion (HASF) module, building a comprehensive anatomical context. The Channel-wise Contrastive Anatomy-Aware Enhancement (CCAE) module further enhances class discrimination and refines segmentation through class-wise channel-level multi-modal contrastive learning. Extensive experiments on the MRSpineSeg and SPIDER datasets demonstrate that ATM-Net significantly outperforms state-of-the-art methods, with consistent improvements regarding class discrimination and segmentation details. For example, ATM-Net achieves Dice of 79.39% and HD95 of 9.91 pixels on SPIDER, significantly outperforming the competitive SpineParseNet by 8.31% and 4.14 pixels, respectively. The integration of anatomical text insights in ATM-Net represents a paradigm innovation in the community, providing clinicians with a powerful tool for precise diagnosis and treatment planning of lumbar spine disorders.

Zitieren

APA

Lian, S., Cai, J., Pan, D., Chen, G., Xu, H., Zhang, F., Fan, G., Pei, J., & Li, S. (2026). Anatomy-Aware Text-Visual Fusion with Dual-Perspective Prompts for Fine-Grained Lumbar Spine Segmentation. International Journal of Computer Vision, 134(5). https://doi.org/10.1007/s11263-026-02852-w

Vancouver

Lian S, Cai J, Pan D, Chen G, Xu H, Zhang F, et al. Anatomy-Aware Text-Visual Fusion with Dual-Perspective Prompts for Fine-Grained Lumbar Spine Segmentation. International Journal of Computer Vision. 2026;134(5). doi:10.1007/s11263-026-02852-w

BibTeX
@article{Lian2026,
  author = {Lian, Sheng and Cai, Jianlong and Pan, Dengfeng and Chen, Guang-Yong and Xu, Hao and Zhang, Fan and Fan, Guodong and Pei, Jialun and Li, Shuo},
  title = {Anatomy-Aware Text-Visual Fusion with Dual-Perspective Prompts for Fine-Grained Lumbar Spine Segmentation},
  journal = {International Journal of Computer Vision},
  year = {2026},
  volume = {134},
  number = {5},
  doi = {10.1007/s11263-026-02852-w}
}
Kostenlos starten →