Digital Staining of High-Definition Fourier Transform Infrared (FT-IR) Images Using Deep Learning

Histological stains, such as hematoxylin and eosin (H&E), are routinely used in clinical diagnosis and research. While these labels offer a high degree of specificity, throughput is limited by the need for multiple samples. Traditional histology stains, such as immunohistochemical labels, also rely only on protein expression and cannot quantify small molecules and metabolites that may aid in diagnosis. Finally, chemical stains and dyes permanently alter the tissue, making downstream analysis impossible. Fourier transform infrared (FT-IR) spectroscopic imaging has shown promise for label-free characterization of important tissue phenotypes and can bypass the need for many chemical labels. Fourier transform infrared classification commonly leverages supervised learning, requiring human annotation that is tedious and prone to errors. One alternative is digital staining, which leverages machine learning to map IR spectra to a corresponding chemical stain. This replaces human annotation with computer-aided alignment. Previous work relies on alignment of adjacent serial tissue sections. Since the tissue samples are not identical at the cellular level, this technique cannot be applied to high-definition FT-IR images. In this paper, we demonstrate that cellular-level mapping can be accomplished using identical samples for both FT-IR and chemical labels. In addition, higher-resolution results can be achieved using a deep convolutional neural network that integrates spatial and spectral features.

[1]  Christian Matthäus,et al.  Label-Free Molecular Imaging of Biological Cells and Tissues by Linear and Nonlinear Raman Spectroscopic Approaches. , 2017, Angewandte Chemie.

[2]  David Mayerich,et al.  High Definition Infrared Spectroscopic Imaging for Lymph Node Histopathology , 2015, PloS one.

[3]  Geoffrey E. Hinton,et al.  Deep Learning , 2015, Nature.

[4]  D. Delpy,et al.  Characterization of the near infrared absorption spectra of cytochrome aa3 and haemoglobin for the non-invasive monitoring of cerebral oxygenation. , 1988, Biochimica et biophysica acta.

[5]  Rohit Bhargava,et al.  Towards a practical Fourier transform infrared chemical imaging protocol for cancer histopathology , 2007, Analytical and bioanalytical chemistry.

[6]  M. Manfait,et al.  Infrared and Raman Imaging for Characterizing Complex Biological Materials: A Comparative Morpho-Spectroscopic Study of Colon Tissue , 2014, Applied spectroscopy.

[7]  Vishal K. Varma,et al.  Infrared spectroscopic imaging: Label-free biochemical analysis of stroma and tissue fibrosis. , 2017, The international journal of biochemistry & cell biology.

[8]  Thomas Brüning,et al.  Marker-free automated histopathological annotation of lung tumour subtypes by FTIR imaging. , 2015, The Analyst.

[9]  Rohit Bhargava,et al.  High-Definition Infrared Spectroscopic Imaging , 2013, Applied spectroscopy.

[10]  Christine Desmedt,et al.  Infrared imaging in breast cancer: automated tissue component recognition and spectral characterization of breast cancer cells as well as the tumor microenvironment. , 2014, The Analyst.

[11]  David Mayerich,et al.  SIproc: an open-source biomedical data processing platform for large hyperspectral images. , 2017, The Analyst.

[12]  Rohit Bhargava,et al.  Using Fourier transform IR spectroscopy to analyze biological materials , 2014, Nature Protocols.

[13]  David Mayerich,et al.  Stain-less staining for computed histopathology. , 2015, Technology.

[14]  Hassan Y. Aboul-Enein,et al.  Vibrational Spectroscopy in Clinical Analysis , 2015 .

[15]  Marian Cholewa,et al.  Application of Raman Spectroscopy and Infrared Spectroscopy in the Identification of Breast Cancer , 2016, Applied spectroscopy.

[16]  Todd C. Hollon,et al.  Rapid intraoperative histology of unprocessed surgical specimens via fibre-laser-based stimulated Raman scattering microscopy , 2017, Nature Biomedical Engineering.

[17]  Rohit Bhargava,et al.  Extracting Knowledge from Chemical Imaging Data Using Computational Algorithms for Digital Cancer Diagnosis , 2015, The Yale journal of biology and medicine.

[18]  K. Pagel,et al.  An infrared spectroscopy approach to follow β-sheet formation in peptide amyloid assemblies. , 2017, Nature chemistry.

[19]  Max Diem,et al.  Statistical analysis of a lung cancer spectral histopathology (SHP) data set. , 2015, The Analyst.

[20]  Pierre Jeannesson,et al.  Development of a memetic clustering algorithm for optimal spectral histology: application to FTIR images of normal human colon. , 2016, The Analyst.

[21]  P Lasch,et al.  Mid-IR microspectroscopic imaging of breast tumor tissue sections. , 2002, Biopolymers.

[22]  S. Hewitt,et al.  Infrared spectroscopic imaging for histopathologic recognition , 2005, Nature Biotechnology.

[23]  Eero P. Simoncelli,et al.  Image quality assessment: from error visibility to structural similarity , 2004, IEEE Transactions on Image Processing.

[24]  Andre Kajdacsy-Balla,et al.  Simultaneous cancer and tumor microenvironment subtyping using confocal infrared microscopy for all-digital molecular histopathology , 2018, Proceedings of the National Academy of Sciences.

[25]  Rohit Bhargava,et al.  High throughput assessment of cells and tissues: Bayesian classification of spectral metrics from infrared vibrational spectroscopic imaging data. , 2006, Biochimica et biophysica acta.

[26]  M. Blades,et al.  Raman Spectroscopy of Blood and Blood Components , 2017, Applied spectroscopy.

[27]  Stimulated Raman scattering microscopy for rapid brain tumor histology , 2017 .

[28]  Georg Bartsch,et al.  Characterization of normal and malignant prostate tissue by Fourier transform infrared microspectroscopy. , 2010, Molecular bioSystems.

[29]  Qingbo Li,et al.  Intraoperative diagnosis of benign and malignant breast tissues by fourier transform infrared spectroscopy and support vector machine classification. , 2015, International journal of clinical and experimental medicine.

[30]  Alex Henderson,et al.  Infrared spectral histopathology using haematoxylin and eosin (H&E) stained glass slides: a major step forward towards clinical translation. , 2017, The Analyst.

[31]  Michael J. Walsh,et al.  Predicting Fibrosis Progression in Renal Transplant Recipients Using Laser-Based Infrared Spectroscopic Imaging , 2018, Scientific Reports.

[32]  D. Lansing Taylor,et al.  Spatial Statistics for Segmenting Histological Structures in H&E Stained Tissue Images , 2017, IEEE Transactions on Medical Imaging.

[33]  David I. Ellis,et al.  Metabolic fingerprinting in disease diagnosis: biomedical applications of infrared and Raman spectroscopy. , 2006, The Analyst.

[34]  Nitish Srivastava,et al.  Dropout: a simple way to prevent neural networks from overfitting , 2014, J. Mach. Learn. Res..

[35]  Ihtesham ur Rehman,et al.  Advances in Fourier transform infrared (FTIR) spectroscopy of biological tissues , 2017 .

[36]  Lin Yang,et al.  Robust Nucleus/Cell Detection and Segmentation in Digital Pathology and Microscopy Images: A Comprehensive Review , 2016, IEEE Reviews in Biomedical Engineering.

[37]  Saurabh Prasad,et al.  Selecting optimal features from Fourier transform infrared spectroscopy for discrete-frequency imaging. , 2018, The Analyst.

[38]  Benjamin Blankertz,et al.  Improving the analysis of near-infrared spectroscopy data with multivariate classification of hemodynamic patterns: a theoretical formulation and validation , 2018, Journal of neural engineering.

[39]  Xiaoou Tang,et al.  Image Super-Resolution Using Deep Convolutional Networks , 2014, IEEE Transactions on Pattern Analysis and Machine Intelligence.