Monocular depth estimation for endoscopic surgery using swin transformer and conditional random fields
Abstract
Monocular depth estimation in endoscopic surgery is challenging due to weak tissue textures, uneven illumination, specular reflections, and frequent smoke or instrument interference. This paper proposes CRFDepth, a pseudo‑supervised framework for endoscopic monocular depth estimation. A Swin Transformer encoder extracts hierarchical global features under weak‑texture conditions. The proposed CRFDepth framework effectively addresses key challenges in endoscopic monocular depth estimation, achieving superior accuracy and robustness. A Pyramid Scene Parsing module aggregates multi‑scale contextual information to improve structural perception in distorted regions. A NeWCRF decoder performs cross‑scale structured inference for boundary refinement and detail recovery. A RangeAwareLoss is introduced to stabilize optimization within the narrow endoscopic depth range. The network is trained with stereo-derived pseudo-ground-truth supervision using binocular image pairs, while only monocular input is required during inference. Experiments were conducted on the da Vinci laparoscopic dataset. CRFDepth achieves 44.27 dB PSNR and 0.9942 SSIM, outperforming baseline methods, and demonstrates strong robustness to occlusion and smoke interference.
Related
- Pseudo-cerebrospinal fluid rhinorrhea following endoscopic endonasal transsphenoidal surgery for PitNET: a case report and literature review
- Blood Eosinophil Count and Lund–Mackay Ethmoid Score as Predictors of Recurrence After Endoscopic Sinus Surgery in CRSwNP
- Subpial dissection technique in endoscopic endonasal surgery for craniopharyngiomas with supradiaphragmatic extension
- Anchored Drill-Assisted Localization for Percutaneous Endoscopic Cervical Keyhole Surgery in the Treatment of Cervical Spondylotic Radiculopathy: A Technical Report and Preliminary Efficacy Evaluation
- Clinical outcomes of uni-portal non-coaxial spinal endoscopic surgery for severe caudally migrated lumbar disc herniation with free fragments
BibTeX
@article{deng2026monocular,
title = {{Monocular depth estimation for endoscopic surgery using swin transformer and conditional random fields}},
author = {Yangyi Deng and Can Yan and Naijun Cheng},
journal = {Scientific Reports},
year = {2026},
month = oct,
doi = {10.1038/s41598-026-73331-y},
url = {https://www.nature.com/articles/s41598-026-73331-y},
note = {CC BY 4.0},
}
RIS
TY - JOUR
TI - Monocular depth estimation for endoscopic surgery using swin transformer and conditional random fields
AU - Yangyi Deng
AU - Can Yan
AU - Naijun Cheng
PY - 2026
DA - 2026/10/01
JO - Scientific Reports
DO - 10.1038/s41598-026-73331-y
UR - https://www.nature.com/articles/s41598-026-73331-y
AB - Monocular depth estimation in endoscopic surgery is challenging due to weak tissue textures, uneven illumination, specular reflections, and frequent smoke or instrument interference. This paper proposes CRFDepth, a pseudo‑supervised framework for endoscopic monocular depth estimation. A Swin Transformer encoder extracts hierarchical global features under weak‑texture conditions. The proposed CRFDepth framework effectively addresses key challenges in endoscopic monocular depth estimation, achieving superior accuracy and robustness. A Pyramid Scene Parsing module aggregates multi‑scale contextual information to improve structural perception in distorted regions. A NeWCRF decoder performs cross‑scale structured inference for boundary refinement and detail recovery. A RangeAwareLoss is introduced to stabilize optimization within the narrow endoscopic depth range. The network is trained with stereo-derived pseudo-ground-truth supervision using binocular image pairs, while only monocular input is required during inference. Experiments were conducted on the da Vinci laparoscopic dataset. CRFDepth achieves 44.27 dB PSNR and 0.9942 SSIM, outperforming baseline methods, and demonstrates strong robustness to occlusion and smoke interference.
SN - 2045-2322
ER -