Laboratory for Spectral Diagnosis, Department of Chemistry and Chemical Biology, Northeastern University, Boston, MA 02115, USA.
Technol Cancer Res Treat. 2011 Apr;10(2):135-44. doi: 10.7785/tcrt.2012.500188.
The detection of micro-metastases and individual metastatic cells in lymph node tissue by spectral methods is summarized. These methods are based on instrument-based acquisition of thousands of infrared spectra of individual tissue pixels from the tissue section, and analysis of the resulting spectral hypercube by multivariate algorithms. The method of infrared image acquisition, followed by multivariate analysis, is henceforth referred to as Spectral Histopathology (SHP). SHP produces pseudo-color images of tissue sections which reveal details that compare very favorably with images collected from hematoxylin/eosin (H and E) stained tissues in that the same tissue structures are detected. However, the infrared results are based on objective and reproducible measurements and do not depend on subjective interpretation. One of the major topics of this paper is the comparison of spectral patterns observed for the same cancer type from different patients. While this is easy in some tissue types, we found it to be difficult in tissues of very different cellularity, or tissue sections that exhibit high levels of inflammatory response. In both cases, spectral quality will be compromised due to confounding effects resulting from scattering effects. The correction of these effects now permits the direct comparison of different patient samples, and paves the way for diagnostic algorithms for cancer detection to be developed.
总结了通过光谱方法检测淋巴结组织中的微转移和单个转移细胞。这些方法基于从组织切片中获取数千个单个组织像素的基于仪器的红外光谱的采集，并通过多元算法分析得到的光谱超立方体。此后，将基于红外图像采集和多元分析的方法称为光谱组织病理学（SHP）。SHP 生成组织切片的伪彩色图像，这些图像显示的细节与从苏木精/伊红（H 和 E）染色组织中收集的图像非常相似，因为可以检测到相同的组织结构。然而，红外结果基于客观和可重复的测量，不依赖于主观解释。本文的主要主题之一是比较来自不同患者的相同癌症类型的光谱模式。虽然在某些组织类型中这很容易，但我们发现对于细胞密度非常不同的组织或表现出高水平炎症反应的组织切片来说，这很困难。在这两种情况下，由于散射效应引起的混杂效应，光谱质量将受到影响。现在，这些效果的校正允许直接比较不同的患者样本，并为开发癌症检测的诊断算法铺平了道路。
