Near real-time nerve visualization using coherent Raman scattering rigid endoscope and deep learning-based image processing for nerve-sparing surgery
Near real-time nerve visualization using coherent Raman scattering rigid endoscope and deep learning-based image processing for nerve-sparing surgery
复制标题
DOI：
10.1117/12.2609483
复制
发表时间：
2022-03
期刊：
影响因子：
--
通讯作者：
N. Yamato;H. Niioka;J. Miyake;M. Hashimoto
中科院分区：
文献类型：
--
作者：
N. Yamato;H. Niioka;J. Miyake;M. Hashimoto
在手术中，应尽可能保留周围神经，以抑制功能障碍，提高术后生活质量。然而，很难将无色、透明和薄的神经与其他组织区分开来。我们开发了一种相干反斯托克斯拉曼散射（汽车）刚性内窥镜，以无标记的方式可视化神经。汽车允许基于分子振动的信息在没有染色的情况下成像。在会议中，我们展示了使用汽车内窥镜和深度学习的近实时神经可视化。我们证明，在1.6秒/图像所采取的图像满足医学图像所需的分割质量。
In surgery, peripheral nerves should be preserved as much as possible to suppress the dysfunction and improve the quality of life after surgery. However, it is difficult to distinguish colorless, transparent, and thin nerves from other tissues. We had developed a coherent anti-Stokes Raman scattering (CARS) rigid endoscope to visualize nerves in a label-free manner. CARS allows for imaging without staining based on the information of molecular vibrations. In the conference, we show near real-time nerve visualization using CARS endoscopy and deep learning. We demonstrate that the image taken at 1.6 s/image satisfies the segmentation quality required for medical images.