Fourier transform infrared spectroscopy (FT-IR) is widely used in the analysis of the chemical composition of biological materials and has the potential to reveal new aspects of the molecular basis of diseases, including different types of cancer. The potential of FT-IR in cancer research lies in its capability of monitoring the biochemical status of cells, which undergo malignant transformation and further examination of spectral features that differentiate normal and cancerous ones using proper mathematical approaches. Such examination can be performed with the use of chemometric tools, such as partial least squares discriminant analysis (PLS-DA) classification and partial least squares regression (PLSR), and proper application of preprocessing methods and their correct sequence is crucial for success. Here, we performed a comparison of several state-of-the-art methods commonly used in infrared biospectroscopy (denoising, baseline correction, and normalization) with the addition of methods not previously used in infrared biospectroscopy classification problems: Mie extinction extended multiplicative signal correction, Eiler’s smoothing, and probabilistic quotient normalization. We compared all of these approaches and their effect on the data structure, classification, and regression capability on experimental FT-IR spectra collected from five different prostate normal and cancerous cell lines. Additionally, we tested the influence of added spectral noise. Overall, we concluded that in the case of the data analyzed here, the biggest impact on data structure and performance of PLS-DA and PLSR was caused by the baseline correction; therefore, much attention should be given, especially to this step of data preprocessing.

Infrared (IR) imaging can be used for fast, accurate and non-destructive pathology recognition of biopsies when supported by machine learning algorithms regardless of the presence of interference effects obscuring the spectra.

In this study, we used polarization-dependent photothermal infrared (OPTIR) microscopy to extend infrared (IR)-based methods to characterize structurally and chemically modified polymer films. Because of the reported relationship with polymer crystallinity and stability, methods capable of monitoring and estimating changes in crystalline regions are paramount for understanding the performance and fate. To develop PTIR methods that could advance the understanding of plastic and polymer degradation, we examined UV-irradiated polyethylene terephthalate (PET) films using simulated photodegradation via a high-energy radiant exposure (SPHERE) weathering device. The well-studied UV-irradiated PET system provided a model to expand the IR characterization capabilities. Previously reported IR methods for determining crystallinity in bulk PET through glycol conformer fractions were combined with advanced structural determination methods to characterize thick and opaque cross-sectioned films using azimuthal angles and Hermans in-plane orientation functions, which extended the sample types available for analysis. To demonstrate the opportunity to decouple signatures from chemical and structural modifications in changing and chemically heterogeneous films, heterogeneity mapping was reported as a function of depth in cross-sectioned films using two-trace two-dimensional (2T2D) correlation analysis coupled with K-means clustering. Faster imaging procedures were also developed to limit possible effects from sample degradation and instrument drift during the acquisition.

Infrared (IR) based histopathology offers a new paradigm in looking at tissues and can provide a complimentary information source for more classical histopathology, which makes it a noteworthy tool given possible clinical application. The goal of this study is to build a robust machine learning model using IR imaging of pancreatic cancer histopathology on a single pixel level. In this article, we report a pancreatic cancer classification model based on data from over 600 biopsies (coming from 250 patients) imaged with IR diffraction-limited spatial resolution. To fully research model’s classification ability, we measured tissues in Standard and High Definition using two optical setups. This forms one of the largest IR datasets analyzed up to now, with almost 700 million spectra of different tissue types. The first classification model, based on six tissue classes, created for comprehensive histopathology achieved AUC values on the pixel (tissue) level above 0.95. We successfully developed a comprehensive histopathology digital staining model for pancreatic tissues based on biochemical information extracted from IR spectra.