High-definition Fourier Transform Infrared (FT-IR) spectroscopic imaging of human tissue sections towards improving pathology

Evaluation of Proteomic and Lipidomic Changes in Aeromonas-Infected Trout Kidney Tissue with the Use of FT-IR Spectroscopy and MALDI Mass Spectrometry Imaging

Aeromonas species are opportunistic bacteria causing a vast spectrum of human diseases, including skin and soft tissue infections, meningitis, endocarditis, peritonitis, gastroenteritis, and finally hemorrhagic septicemia. The aim of our research was to indicate the molecular alterations in proteins and lipids profiles resulting from Aeromonas sobria and A. salmonicida subsp. salmonicida infection in trout kidney tissue samples. We successfully applied FT-IR (Fourier transform infrared) spectroscopy and MALDI-MSI (matrix-assisted laser desorption/ionization mass spectrometry imaging) to monitor changes in the structure and compositions of lipids, secondary conformation of proteins, and provide useful information concerning disease progression. Our findings indicate that the following spectral bands’ absorbance ratios (spectral biomarkers) can be used to discriminate healthy tissue from pathologically altered tissue, for example, lipids (CH2/CH3), amide I/amide II, amide I/CH2 and amide I/CH3. Spectral data obtained from 10 single measurements of each specimen indicate numerous abnormalities concerning proteins, lipids, and phospholipids induced by Aeromonas infection, suggesting significant disruption of the cell membranes. Moreover, the increase in the content of lysolipids such as lysophosphosphatidylcholine was observed. The results of this study suggest the application of both methods MALDI-MSI and FT-IR as accurate methods for profiling biomolecules and identifying biochemical changes in kidney tissue during the progression of Aeromonas infection.

Mid-Infrared Imaging Characterization to Differentiate Lung Cancer Subtypes

Introduction: Lung cancer is the most common malignancy worldwide. Squamous cell carcinoma (SQ) and adenocarcinoma (LUAD) are the two most frequent histological subtypes. Small cell carcinoma (SCLC) subtype has the worst prognosis. Differential diagnosis is essential for proper oncological treatment. Life science associated mid- and near-infrared based microscopic techniques have been developed exponentially, especially in the past decade. Vibrational spectroscopy is a potential non-destructive approach to investigate malignancies.

Aims: Our goal was to differentiate lung cancer subtypes by their label-free mid-infrared spectra using supervised multivariate analyses.

Material and Methods: Formalin-fixed paraffin-embedded (FFPE) samples were selected from the archives. Three subtypes were selected for each group: 10-10 cases SQ, LUAD and SCLC. 2 μm thick sections were cut and laid on aluminium coated glass slides. Transflection optical setup was applied on Perkin-Elmer infrared microscope. 250 × 600 μm areas were imaged and the so-called mid-infrared fingerprint region (1800-648cm⁻¹) was further analysed with linear discriminant analysis (LDA) and support vector machine (SVM) methods.

Results: Both “patient-based” and “pixel-based” approaches were examined. Patient-based analysis by using 3 LDA models and 2 SVM models resulted in different separations. The higher the cut-off value the lower is the accuracy. The linear C-support vector classification (C-SVC) SVM resulted in the best (100%) accuracy for the three subtypes using a 50% cut-off value. The pixel-based analysis gave, similarly, the linear C-SVC SVM model to be the most efficient in the statistical indicators (SQ sensitivity 81.65%, LUAD sensitivity 82.89% and SCLC sensitivity 88.89%). The spectra cut-off, the kernel function and the algorithm function influence the accuracy.

Conclusion: Mid-Infrared imaging could be used to differentiate FFPE lung cancer subtypes. Supervised multivariate tools are promising to accurately separate lung tumor subtypes. The long-term perspective is to develop a spectroscopy-based diagnostic tool, revolutionizing medical differential diagnostics, especially cancer identification.

Statistical Considerations and Tools to Improve Histopathologic Protocols with Spectroscopic Imaging

Advances in infrared (IR) spectroscopic imaging instrumentation and data science now present unique opportunities for large validation studies of the concept of histopathology using spectral data. In this study, we examine the discrimination potential of IR metrics for different histologic classes to estimate the sample size needed for designing validation studies to achieve a given statistical power and statistical significance. Next, we present an automated annotation transfer tool that can allow large-scale training/validation, overcoming the limitations of sparse ground truth data with current manual approaches by providing a tool to transfer pathologist annotations from stained images to IR images across diagnostic categories. Finally, the results of a combination of supervised and unsupervised analysis provide a scheme to identify diagnostic groups/patterns and isolating pure chemical pixels for each class to better train complex histopathological models. Together, these methods provide essential tools to take advantage of the emerging capabilities to record and utilize large spectroscopic imaging datasets.

High-resolution infrared microspectroscopic characterization of cartilage cell microenvironment

The lateral resolution of infrared spectroscopy has been inadequate for accurate biochemical characterization of the cell microenvironment, a region regulating biochemical and biomechanical signals to cells. In this study, we demonstrate the capacity of a high-resolution Fourier transform infrared microspectroscopy (HR-FTIR-MS) to characterize the collagen content of this region. Specifically, we focus on the collagen content in the cartilage cell (chondrocyte) microenvironment of healthy and osteoarthritic (OA) cartilage. Human tibial cartilage samples (N = 28) were harvested from 7 cadaveric donors and graded for OA severity (healthy, early OA, advanced OA). HR-FTIR-MS was used to analyze the collagen content of the chondrocyte microenvironment of five distinct zones across the tissue depth. HR-FTIR-MS successfully showed collagen content distribution across chondrocytes and their environment. In zones 2 and 3 (10 - 50% of the tissue thickness), we observed that collagen content was smaller (P < 0.05) in early OA compared to the healthy tissue in the vicinity of cells (pericellular region). The collagen content loss was extended to the extracellular matrix in advanced OA tissue. No significant differences in the collagen content of the chondrocyte microenvironment were observed between the groups in the most superficial (0-10%) and deep zones (50-100%). HR-FTIR-MS revealed collagen loss in the early OA cartilage pericellular region before detectable changes in the extracellular matrix in advanced OA. HR-FTIR-MS-based compositional assessment enables a better understanding of OA-related changes in tissues. This technique can be used to identify new disease mechanisms enabling better intervention strategies. STATEMENT OF SIGNIFICANCE: Osteoarthritis (OA) is the most common degenerative joint disease causing pain and disability. While significant progress has been made in OA research, OA pathogenesis is still poorly understood and current OA treatments are mainly palliative. This study demonstrates that high-resolution FTIR microspectroscopy (HR-FTIR-MS) can characterize OA-induced compositional changes in the cell microenvironment (pericellular matrix) during the early disease stages before tissue changes in the extracellular matrix become apparent. This technique may further enable the identification of new OA mechanisms and improve our current understanding of OA pathogenesis, thus, enabling the development of better treatment methods.

Design Considerations for Discrete Frequency Infrared Microscopy Systems

Discrete frequency infrared chemical imaging is transforming the practice of microspectroscopy by enabling a diversity of instrumentation and new measurement capabilities. While a variety of hardware implementations have been realized, design considerations that are unique to infrared (IR) microscopes have not yet been compiled in literature. Here, we describe the evolution of IR microscopes, provide rationales for design choices, and catalog some major considerations for each of the optical components in an imaging system. We analyze design choices that use these components to optimize performance, under their particular constraints, while providing illustrative examples. We then summarize a framework to assess the factors that determine an instrument's performance mathematically. Finally, we provide a validation approach by enumerating performance metrics that can be used to evaluate the capabilities of imaging systems or suitability for specific intended applications. Together, the presented concepts and examples should aid in understanding available instrument configurations, while guiding innovations in design of the next generation of IR chemical imaging spectrometers.

Mid-Infrared Imaging Is Able to Characterize and Separate Cancer Cell Lines

Malignancies are still responsible for a large share of lethalities. Macroscopical evaluation of the surgical resection margins is uncertain. Big data based imaging approaches have emerged in the recent decade (mass spectrometry, two-photon microscopy, infrared and Raman spectroscopy). Indocianine green labelled MS is the most common approach, however, label free mid-infrared imaging is more promising for future practical application. We aimed to identify and separate different transformed (A-375, HT-29) and non-transformed (CCD986SK) cell lines by a label-free infrared spectroscopy method. Our approach applied a novel set-up for label-free mid-infrared range classification method. Transflection spectroscopy was used on aluminium coated glass slides. Both whole range spectra (4000-648 cm^-1) and hypersensitive fingerprint regions (1800-648 cm^-1) were tested on the imaged areas of cell lines fixed in ethanol. Non-cell spectra were possible to be excluded based on mean transmission values being above 90%. Feasibility of a mean transmission based spectra filtering method with principal component analysis and linear discriminant analysis was shown to separate cell lines representing different tissue types. Fingerprint region resulted the best separation of cell lines spectra with accuracy of 99.84% at 70-75 mean transmittance range. Our approach in vitro was able to separate unique cell lines representing different tissues of origin. Proper data handling and spectra processing are key steps to achieve the adaptation of this dye-free technique for intraoperative surgery. Further studies are urgently needed to test this novel, marker-free approach.

Biochemical similarity between cultured chondrocytes and in situ chondrocytes by chemometric analysis from FTIR microspectroscopy

Background aims

Fourier Transform Infrared Micro-spectroscopy (FTIRM) is an emerging tool that obtains images with biochemical information of samples that are too small to be chemically analyzed by conventional Fourier transform infrared (FTIR) spectroscopy techniques. So, the central objective of this project was to study the biochemical similarity between articular and cultured chondrocytes by chemometric analysis from FTIRM.

Methods

Nine samples of knee articular cartilage were obtained; each sample was divided into two fragments, one portion was used for FTIRM characterization in situ, and from another part, chondrocytes were obtained to be cultured (in vitro), which were subjected to an FTIRM to characterize their biomolecular components. The FTIRM spectra were normalized, and the second derivative was calculated. From these data, principal component analysis (PCA) and a chemometric comparison between in situ and cultured chondrocytes were carried out. Finally, the biochemical mapping was conducted obtaining micro-FTIR imaging.

Results

FTIRM spectra of in situ and in vitro chondrocytes were obtained, and different biomolecules were detected, highlighting lipids, proteins, glycosaminoglycans, collagen, and aggrecan. Despite slight differences in the FTIR spectra, the PCA proved the organic similarity between in situ chondrocytes and cultured chondrocytes, which was also observed in the analysis of the ratios related to the degradation of the articular cartilage and collagen. In the same way, the ability of the FTIRM to characterize the molecular biodistribution was demonstrated.

Conclusion

The biochemical composition and biodistribution analysis using FTIRM have been useful for comparing cultured chondrocytes and in situ chondrocytes.

Reliability of Histopathology for the Early Recognition of Fibrosis in Traction Alopecia: Correlation with Clinical Severity

Traction alopecia (TA) is hair loss caused by prolonged pulling or repetitive tension on scalp hair; it belongs to the biphasic group of primary alopecia. It is non-scarring, typically with preservation of follicular stem cells and the potential for regrowth of early lesions especially if traction hairstyles are stopped. However, the alopecia may become permanent (scarring) and fail to respond to treatment if the traction is excessive and prolonged. Hence, the ability to detect fibrosis early in these lesions could predict patients who respond to treatment. Histopathological diagnosis based on scalp biopsies has been used as a gold standard to delineate various forms of non-scarring alopecia and to differentiate them from scarring ones. However, due to potential discrepant reporting as a result of the type of biopsy, method of sectioning, and site of biopsy, histopathology often tends to be unreliable for the early recognition of fibrosis in TA. In this study, 45 patients were assessed using the marginal TA severity scoring system, and their biopsies (both longitudinal and transverse sections) were systematically assessed by three dermatopathologists, the aim being to correlate histopathological findings with clinical staging. Intraclass correlation coefficients were used to determine the level of agreement between the assessors. We found poor agreement of the identification and grading of perifollicular and interfollicular fibrosis (0.55 [0.23–0.75] and 0.01 [2.20–0.41], respectively), and no correlation could be drawn with the clinical severity score. Better methods of diagnosis are needed for grading and for recognition of early fibrosis in TA.

Macromolecular Characterization of Swordfish Oocytes by FTIR Imaging Spectroscopy

During folliculogenesis, primary oocytes of teleosts grow by several orders of magnitude by-self synthesizing proteins and mRNA, or sequestering from blood specific macromolecular components, such as fatty acids and vitellogenin. All these materials are stored into cortical alveoli, yolk globules or oil droplets during oocyte development. The proper synthesis, storage and displacement of these macromolecular components inside the oocyte play a key role for a successful fertilization process and for the subsequently correct embryo development. In this study, for the first time, the FTIR Imaging (FTIRI) spectroscopy has been applied to characterize the chemical building blocks of several cellular components of swordfish oocytes at different developmental stages. In particular, the spectral features of previtellogenic (PV), vitellogenic (VTG), mature (M) and atretic (A) follicles as well as and of cortical alveoli (CA), yolk vesicles (YV), oil droplets (OD) and Zona Radiata (ZR) have been outlined, providing new insights in terms of composition and topographical distribution of macromolecules of biological interest such as lipids, proteins, carbohydrates and phosphates. The macromolecular characterization of swordfish oocytes at different developmental stages represents a starting point and a useful tool for the assessment of swordfish egg quality caught in different conditions, such as periods of the year or different fishing area.

A fully automated, faster noise rejection approach to increasing the analytical capability of chemical imaging for digital histopathology

Chemical hyperspectral imaging (HSI) data is naturally high dimensional and large. There are thus inherent manual trade-offs in acquisition time, and the quality of data. Minimum Noise Fraction (MNF) developed by Green et al. [1] has been extensively studied as a method for noise removal in HSI data. It too, however entails a manual speed-accuracy trade-off, namely the process of manually selecting the relevant bands in the MNF space. This process currently takes roughly around a month’s time for acquiring and pre-processing an entire TMA with acceptable signal to noise ratio. We present three approaches termed ‘Fast MNF’, ‘Approx MNF’ and ‘Rand MNF’ where the computational time of the algorithm is reduced, as well as the entire process of band selection is fully automated. This automated approach is shown to perform at the same level of accuracy as MNF with now large speedup factors, resulting in the same task to be accomplished in hours. The different approximations produced by the three algorithms, show the reconstruction accuracy vs storage (50×) and runtime speed (60×) trade-off. We apply the approach for automating the denoising of different tissue histology samples, in which the accuracy of classification (differentiating between the different histologic and pathologic classes) strongly depends on the SNR (signal to noise ratio) of recovered data. Therefore, we also compare the effect of the proposed denoising algorithms on classification accuracy. Since denoising HSI data is done unsupervised, we also use a metric that assesses the quality of denoising in the image domain between the noisy and denoised image in the absence of ground truth.

Emerging Themes in Image Informatics and Molecular Analysis for Digital Pathology

Pathology is essential for research in disease and development, as well as for clinical decision making. For more than 100 years, pathology practice has involved analyzing images of stained, thin tissue sections by a trained human using an optical microscope. Technological advances are now driving major changes in this paradigm toward digital pathology (DP). The digital transformation of pathology goes beyond recording, archiving, and retrieving images, providing new computational tools to inform better decision making for precision medicine. First, we discuss some emerging innovations in both computational image analytics and imaging instrumentation in DP. Second, we discuss molecular contrast in pathology. Molecular DP has traditionally been an extension of pathology with molecularly specific dyes. Label-free, spectroscopic images are rapidly emerging as another important information source, and we describe the benefits and potential of this evolution. Third, we describe multimodal DP, which is enabled by computational algorithms and combines the best characteristics of structural and molecular pathology. Finally, we provide examples of application areas in telepathology, education, and precision medicine. We conclude by discussing challenges and emerging opportunities in this area.

Appropriate Tensile Mode and Timing of Applying Tension to Promote Tendon Gel Regeneration

"Tendon gel" secreted from a parent tendon is regenerated for tendon repair by applying tension. However, the details of the tensile stimulus have not been clarified. This study aimed to evaluate an appropriate tensile stimulus mode and the optimal timing of applying tension to promote tendon gel regeneration. Tendon gel was prepared using a film model method in mice and was preserved in vivo for 3, 5, and 10 days. Unlike tendon gel on day 3 or day 5, a fibrous structure developed in the tendon gel on day 10 when tension was applied. Infrared spectroscopy revealed that characteristic peaks appearing for the tendon gel on days 3 and 5 disappeared on day 10. Disappearance of the peaks indicated maturity of the tendon gel, and it showed the optimal timing for tension application to the tendon gel. The effect of tensile load on tendon gel preserved for 10 days was investigated using a tensile test, a creep test, or a cycle test. In the tensile test, tendon gel was elongated into a thin cord of collagen fibers with an increase in stress, and the maximum diameter of the collagen fiber was approximately 50 times larger than that in the normal Achilles tendon of mice. The results suggest that the diameter of the oriented collagen fiber is controllable by adjusting the applied load and the time in mature tendon gel.

Development of a practical spatial-spectral analysis protocol for breast histopathology using Fourier transform infrared spectroscopic imaging

Breast cancer screening provides sensitive tumor identification, but low specificity implies that a vast majority of biopsies are not ultimately diagnosed as cancer. Automated techniques to evaluate biopsies can prevent errors, reduce pathologist workload and provide objective analysis. Fourier transform infrared (FT-IR) spectroscopic imaging provides both molecular signatures and spatial information that may be applicable for pathology. Here, we utilize both the spectral and spatial information to develop a combined classifier that provides rapid tissue assessment. First, we evaluated the potential of IR imaging to provide a diagnosis using spectral data alone. While highly accurate histologic [epithelium, stroma] recognition could be achieved, the same was not possible for disease [cancer, no-cancer] due to the diversity of spectral signals. Hence, we employed spatial data, developing and evaluating increasingly complex models, to detect cancers. Sub-mm tumors could be very confidently predicted as indicated by the quantitative measurement of accuracy via receiver operating characteristic (ROC) curve analyses. The developed protocol was validated with a small set and statistical performance used to develop a model that predicts study design for a large scale, definitive validation. The results of evaluation on different instruments, at higher noise levels, under a coarser spectral resolution and two sampling modes [transmission and transflection], indicate that the protocol is highly accurate under a variety of conditions. The study paves the way to validating IR imaging for rapid breast tumor detection, its statistical validation and potential directions for optimization of the speed and sampling for clinical deployment.