Affiliations: Leibniz Institute of Photonic Technology, Jena, Germany | Department of Otorhinolaryngology, University Hospital Jena, Jena, Germany | Institute of Physical Chemistry and Abbe Center of Photonics, University Jena, Jena, Germany
Note: [] Corresponding author: Christoph Krafft, Leibniz Institute of Photonic Technology, Albert-Einstein Str. 9, 07745 Jena, Germany. E-mail: [email protected]
Abstract: The advent of FTIR microscopic spectrometers with focal plane array detectors enabled rapid image acquisition with diffraction limited lateral resolution. The field of view depends on the magnification and the detector size. FTIR images of large samples can be collected in the so called mosaic mode by stitching individual images together. If the mosaic is composed of hundreds of images, the total acquisition time and the data size will increase considerably. One computational and two optical options are compared to reduce both acquisition time and data size. First, the 2× field expansion optic increases the measured sample area fourfold. Second, using a 4× objective instead of the standard 15× objective increases the area covered by a single image by a factor of 11. Third, pixel binning averages neighboring pixels at the expense of lateral resolution. All options are demonstrated in a case study of a thin section of laryngeal carcinoma encompassing normal tissue, inflammation, connective tissue, dysplasia, carcinoma and blood. Data analysis is described using the toolbox hyperSpec operating under the R environment and complemented by parallel computing functions. A classification model that was trained with low magnification data in the range from 1200 to 1800 cm−1 gave similar results for higher magnification data. Restrictions occurred for microscopic features smaller than the detector pixel size and for biomarkers below 1200 cm−1 due to signal attenuation of the 4× objective lenses. FTIR imaging mosaic strategies of other groups and the emerging use of quantum cascade lasers for IR imaging are discussed.
Keywords: Infrared spectroscopy, infrared microscopy, tissue diagnosis, chemometric data analysis
