Using optical techniques for tissue diagnostics (so-called ‘optical biopsy’) has been a subject of extensive research for many years. Various groups have been exploring different spectral and/or imaging modalities (e.g. diffuse reflectance spectroscopy, autofluorescence, Raman spectroscopy, optical coherence tomography (OCT), polarized light microscopy, etc.) for biomedical applications. In this paper, we report on using multi-wavelength imaging Mueller polarimetry combined with an appropriated image post-processing for the detection of tissue malignancy. We investigate a possibility of complementary analysis of Mueller matrix images obtained for turbid tissue-like scattering phantoms and excised human normal and cancerous colorectal tissue samples embedded in paraffin. Combined application of correlation, fractal and statistical analysis was employed to assess quantitatively the polarization-inhomogeneous scattered fields observed at the surface of tissue samples. The combined analysis of the polarimetric images of paraffin-embedded tissue blocks has proved to be an efficient tool for the unambiguous detection of tissue malignant transformation. A fractal structure was clearly observed at spatial distributions of depolarization of light scattered in healthy tissues in a visible range of spectrum, while corresponding distributions for cancerous tissues did not show such dependence. We demonstrate that paraffin does not destroy a fractal structure of spatial distribution of depolarization. Thus, the loss of fractality in spatial distributions of depolarization for cancerous tissue is related to the structural changes in the tissue sample induced by cancer itself and, therefore, may serve as a marker of the disease. The obtained results emphasize that a combined use of statistical, correlation and fractal analysis for the Mueller-matrix image post-processing is an effective approach for an assessment of variations of optical properties in turbid tissue-like scattering media and biological tissues, with a high potential to be transferred to clinical practice for screening cancerous tissue samples.

中文翻译：

---

光学各向异性高散射生物组织的Mueller矩阵图像的互补分析

使用光学技术进行组织诊断（所谓的“光学活检”）多年来一直是广泛研究的主题。各个小组一直在探索用于生物医学应用的不同光谱和/或成像方式（例如，漫反射光谱，自发荧光，拉曼光谱，光学相干断层扫描（OCT），偏振光显微镜等）。在本文中，我们报告了使用多波长成像Mueller旋光仪结合适当的图像后处理来检测组织恶性肿瘤。我们调查的可能性的Mueller矩阵图像的补充分析的可能性为浑浊的组织状散射体模和石蜡包埋的切除的人类正常和癌性结直肠组织样本。结合应用相关，分形和统计分析用于定量评估在组织样品表面观察到的偏振不均匀散射场。结合石蜡包埋的组织块的极化图像分析已被证明是一种有效检测组织恶性转化的有效工具。在可见光谱范围内，在健康组织中散射的光的去极化空间分布中清晰地观察到分形结构，而癌组织的相应分布则没有这种依赖性。我们证明石蜡不会破坏去极化的空间分布的分形结构。从而，癌组织去极化空间分布中的分形性损失与癌症本身引起的组织样本结构变化有关，因此可以作为疾病的标志物。获得的结果强调，对Mueller-matrix图像后处理进行统计，相关和分形分析的组合使用是评估浑浊的类组织散射介质和生物组织中光学性质变化的有效方法，具有很高的可能被转移到临床实践中以筛查癌变组织样品。