文献速递鼻科

基于计算机视觉的视觉引导手术导航用于术中动态影像更新

Vision-Guided Surgical Navigation Using Computer Vision for Dynamic Intraoperative Imaging Updates.
PubMed 文献 · Int Forum Allergy Rhinol · 2025-08-22

本研究在三个3D打印模型(共6侧)上模拟双侧鼻内镜鼻窦手术,利用神经辐射场(NeRF)从术后内镜视频生成3D重建并与术前CT配准,通过算法减除切除区域生成数字化更新CT。更新CT与金标准术后CT的平均Hausdorff距离为0.27±0.076 mm,Dice相似系数为0.93±0.012,Bland-Altman分析显示体积平均差为0.40 cm3(95%一致性界限0.04–0.76 cm3)。作者认为计算机视觉可无需外部追踪即可从单目内镜视频生成高精度CT更新。

要点

  • 在3D打印模型上,NeRF生成的更新CT与金标准术后CT表面配准达亚毫米级。
  • Dice相似系数0.93±0.012,提示更新CT与金标准CT空间重叠优异。
  • Bland-Altman分析显示更新CT体积平均低估0.40 cm3,95%一致性界限0.04–0.76 cm3。

核心结论

  • 计算机视觉可通过单目内镜视频生成高精度CT更新,无需外部追踪即可实现术中动态影像。
  • 作者认为该软件驱动工具有潜力在下一代导航平台中提升鼻内镜鼻窦手术的切除完整性。

研究设计与样本

  • 在三个3D打印模型(共6侧)上实施双侧鼻内镜鼻窦手术,术后内镜视频经定制NeRF流程生成3D重建并与术前CT配准。
  • 通过算法减除切除区域生成数字化更新CT,进行体积分割后与金标准术后CT比较,采用Hausdorff距离、Dice相似系数和Bland-Altman分析评估准确性。

主要发现

  • 更新CT与金标准术后CT的平均Hausdorff距离为0.27±0.076 mm,第95百分位Hausdorff距离为0.82±0.165 mm,提示亚毫米级表面配准。
  • Dice相似系数为0.93±0.012,数值大于0.9提示空间重叠优异。
  • Bland-Altman分析显示更新CT较金标准CT体积平均低估0.40 cm3,95%一致性界限为0.04–0.76 cm3,所有样本均落在可接受的变异范围内。

临床怎么用

  • 原文未据此给出临床推荐。
  • 作者认为该软件驱动工具有潜力在下一代导航平台中提升鼻内镜鼻窦手术的切除完整性。

局限性

  • 原文未报告局限性。

以上解读依据论文摘要(全文未开放获取或许可不允许翻译转载)。

原文

BACKGROUND: Residual disease after endoscopic sinus surgery (ESS) contributes to poor outcomes and revision surgery. Image-guided surgery systems cannot dynamically reflect intraoperative changes. We propose a sensorless, video-based method for intraoperative CT updating using neural radiance fields (NeRF), a deep learning algorithm used to create 3D surgical field reconstructions. METHODS: Bilateral ESS was performed on three 3D-printed models (n = 6 sides). Postoperative endoscopic videos were processed through a custom NeRF pipeline to generate 3D reconstructions, which were co-registered to preoperative CT scans. Digitally updated CT models were created through algorithmic subtraction of resected regions, then volumetrically segmented, and compared to ground-truth postoperative CT. Accuracy was assessed using Hausdorff distance (surface alignment), Dice similarity coefficient (DSC) (volumetric overlap), and Bland‒Altman analysis (BAA) (statistical agreement). RESULTS: Comparison of the updated CT and the ground-truth postoperative CT indicated an average Hausdorff distance of 0.27 ± 0.076 mm and a 95th percentile Hausdorff distance of 0.82 ± 0.165 mm, indicating sub-millimeter surface alignment. The DSC was 0.93 ± 0.012 with values >0.9 suggestive of excellent spatial overlap. BAA indicated modest underestimation of volume on the updated CT versus ground-truth CT with a mean difference in volumes of 0.40 cm3 with 95% limits of agreement of 0.04‒0.76 cm3 indicating that all samples fell within acceptable bounds of variability. CONCLUSIONS: Computer vision can enable dynamic intraoperative imaging by generating highly accurate CT updates from monocular endoscopic video without external tracking. By directly visualizing resection progress, this software-driven tool has the potential to enhance surgical completeness in ESS for next-generation navigation platforms.

来源:Int Forum Allergy Rhinol,查看原文。版权归原作者所有。

引用

  1. Ruthberg J, Gunderson N, Chen P, et al. Vision-Guided Surgical Navigation Using Computer Vision for Dynamic Intraoperative Imaging Updates. Int Forum Allergy Rhinol. 2025. PMID: 40847589. doi:10.1002/alr.70000

供执业医师工作参考,不替代临床判断;数值与分级以指南原文为准。