IEEE Trans Med Imaging. 2019 Jan;38(1):261-268. doi: 10.1109/TMI.2018.2861570. Epub 2018 Jul 31.
The attenuation coefficient is a relevant biomarker for many diagnostic medical applications. Recently, the Depth-Resolved Confocal (DRC) technique was developed to automatically estimate the attenuation coefficients from Optical Coherence Tomography (OCT) data with pixel-level resolution. However, DRC requires that the confocal function parameters (i.e., focal plane location and apparent Rayleigh range) be known a priori. In this paper, we present the autoConfocal algorithm: a simple, automatic method for estimating those parameters directly from OCT imagery when the focal plane is within the sample. We present autoConfocal+DRC results on phantom data, ex-vivo biological tissue data, and in-vivo clinical data.
衰减系数是许多诊断医学应用的相关生物标志物。最近,深度分辨共焦(DRC)技术被开发出来,可从具有像素级分辨率的光相干断层扫描(OCT)数据中自动估计衰减系数。然而,DRC 要求共焦函数参数(即焦平面位置和表观瑞利范围)预先已知。在本文中,我们提出了 autoConfocal 算法:一种简单、自动的方法,可在焦平面位于样品内时直接从 OCT 图像估计这些参数。我们在体模数据、离体生物组织数据和体内临床数据上展示了 autoConfocal+DRC 的结果。