Gorpas Dimitris, Ma Dinglong, Bec Julien, Yankelevich Diego R, Marcu Laura
IEEE Trans Med Imaging. 2016 Aug;35(8):1802-11. doi: 10.1109/TMI.2016.2530621. Epub 2016 Feb 15.
Fiber based fluorescence lifetime imaging has shown great potential for intraoperative diagnosis and guidance of surgical procedures. Here we describe a novel method addressing a significant challenge for the practical implementation of this technique, i.e., the real-time display of the quantified biochemical or functional tissue properties superimposed on the interrogated area. Specifically, an aiming beam (450 nm) generated by a continuous-wave laser beam was merged with the pulsed fluorescence excitation light in a single delivery/collection fiber and then imaged and segmented using a color-based algorithm. We demonstrate that this approach enables continuous delineation of the interrogated location and dynamic augmentation of the acquired frames with the corresponding fluorescence decay parameters. The method was evaluated on a fluorescence phantom and fresh tissue samples. Current results demonstrate that 34 frames per second can be achieved for augmenting videos of 640 × 512 pixels resolution. Also we show that the spatial resolution of the fluorescence lifetime map depends on the tissue optical properties, the scanning speed, and the frame rate. The dice similarity coefficient between the fluorescence phantom and the reconstructed maps was estimated to be as high as 93%. The reported method could become a valuable tool for augmenting the surgeon's field of view with diagnostic information derived from the analysis of fluorescence lifetime data in real-time using handheld, automated, or endoscopic scanning systems. Current method provides also a means for maintaining the tissue light exposure within safety limits. This study provides a framework for using an aiming beam with other point spectroscopy applications.
基于光纤的荧光寿命成像在手术过程的术中诊断和引导方面已显示出巨大潜力。在此，我们描述了一种新颖的方法，该方法解决了这项技术实际应用中的一个重大挑战，即叠加在被询问区域上的生物化学或功能组织特性的定量实时显示。具体而言，由连续波激光束产生的瞄准光束（450纳米）与脉冲荧光激发光在单根传输/收集光纤中合并，然后使用基于颜色的算法进行成像和分割。我们证明，这种方法能够持续描绘被询问位置，并利用相应的荧光衰减参数对采集的帧进行动态增强。该方法在荧光模型和新鲜组织样本上进行了评估。当前结果表明，对于分辨率为640×512像素的增强视频，每秒可实现34帧。我们还表明，荧光寿命图的空间分辨率取决于组织光学特性、扫描速度和帧率。荧光模型与重建图之间的骰子相似系数估计高达93%。所报道的方法可能成为一种有价值的工具，可利用手持、自动或内窥镜扫描系统实时分析荧光寿命数据得出的诊断信息来扩展外科医生的视野。当前方法还提供了一种将组织光暴露保持在安全限度内的手段。本研究为将瞄准光束用于其他点光谱应用提供了一个框架。
