Liao Haolin, Mitten David J, Knox Wayne H
The Institute of Optics, University of Rochester, Rochester, NY 14627, USA.
Department of Orthopaedics, University of Rochester, Rochester, NY 14627, USA.
Biomed Opt Express. 2025 Jan 27;16(2):679-689. doi: 10.1364/BOE.545146. eCollection 2025 Feb 1.
In this work, we present a real-time method to aid intraoperative peripheral nerve identification. Using LEDs as the light sources, the device contains a driving motor that rotates a pair of orthogonally oriented linear polarizers. By performing lock-in processing to frames taken under the rotating crossed polarization imaging (RXPI) system, the AC components of the periodic intensity signal of chicken tissues are acquired and compared. With an area under the curve (AUC) of 93%, the chicken sciatic nerve is distinct for automatic identification. In both chicken thigh and cadaver arm models, the contrasts of nerve tissues are successfully enhanced in the lock-in processed output image. Real-time automatic nerve masking is successfully demonstrated in the chicken model using a portable prototype weighing 525 g. In conclusion, the RXPI system with lock-in processing methods can potentially serve as an intraoperative nerve identification aid.
在这项工作中,我们提出了一种辅助术中识别周围神经的实时方法。该设备以发光二极管作为光源,包含一个驱动电机,用于旋转一对正交定向的线性偏振器。通过对在旋转交叉偏振成像(RXPI)系统下拍摄的帧进行锁相处理,获取并比较鸡组织周期性强度信号的交流分量。鸡坐骨神经的曲线下面积(AUC)为93%,可实现自动识别。在鸡大腿和尸体手臂模型中,锁相处理后的输出图像中神经组织的对比度均得到成功增强。使用一个重525克的便携式原型在鸡模型中成功演示了实时自动神经掩蔽。总之,采用锁相处理方法的RXPI系统有潜力作为术中神经识别的辅助工具。