The evolutionary behavior is examined for heterogeneously distributed hyperspectral images of a simulated biological tissue sample comprising lipid-like and protein-like components during the aging process. Taking a simple planar average of a spectral image loses useful information about the spatially resolved nature of the data. In contrast, multivariate curve resolution (MCR) analysis of a spectral image at a given stage of aging produces a set of loadings of major component groups. Each loading represents the combined spectral contributions of a mixture of similar but not identical constituents (i.e., lipid-like and protein-like components). Temporal analysis of individual component groups using two-dimensional correlation spectroscopy (2D-COS) and MCR provides much-streamlined results without interferences from the overlapped contributions. Grouping of data into separate components also allows for the effective comparison of the parallel processes of lipid oxidation and protein denaturation involving a number of constituents using the heterocomponent 2D-COS analysis. The complex interplays of lipid constituents and protein secondary structures during the tissue aging process are unambiguously highlighted. The possibility of extending this approach to a much more general form of applications using a moving window analysis is also discussed.

中文翻译：

---

演化高光谱图像的二维相关光谱 (2D-COS) 分析

检查老化过程中包含类脂和类蛋白成分的模拟生物组织样本的异质分布高光谱图像的进化行为。对光谱图像进行简单的平面平均会丢失有关数据空间分辨性质的有用信息。相比之下，在给定的老化阶段对光谱图像进行多元曲线分辨率 (MCR) 分析会产生一组主要成分组的负载。每个负载代表相似但不相同的成分（即类脂质和类蛋白质成分）的混合物的组合光谱贡献。使用二维相关光谱 (2D-COS) 和 MCR 对各个组分组进行时间分析可提供更加简化的结果，而不会受到重叠贡献的干扰。将数据分组为单独的成分还可以使用异质成分 2D-COS 分析来有效比较涉及多种成分的脂质氧化和蛋白质变性的并行过程。明确强调了组织老化过程中脂质成分和蛋白质二级结构的复杂相互作用。还讨论了使用移动窗口分析将此方法扩展到更通用的应用程序形式的可能性。