FTIR Microscopic Imaging of Large Samples with 4x and 15x Infrared Objectives: A Case Study of a Carcinoma Tissue Section

Significance of the topic

The ability to map biochemical composition across large tissue sections by FTIR microscopic imaging supports rapid, label-free analysis in biomedical research and quality control. Expanding the field of view without sacrificing spectral quality is crucial for efficient screening and detailed histopathological assessment.

Objectives and Study Overview

This case study compares two infrared objectives (4x and 15x) in an Agilent FTIR microscope for imaging a laryngeal carcinoma tissue section. The goal is to evaluate field of view, spatial resolution, data volume, acquisition time, and tissue classification performance.

Methodology and Instrumentation

An Agilent Cary 620 FTIR microscope with a 64×64 FPA detector was coupled to an Agilent Cary 670 FTIR spectrometer. Measurements employed 4 cm⁻¹ spectral resolution, 16 co-added sample scans, 32 background scans, and an under-sampling setting of 4, covering up to 3975 cm⁻¹ with a low-pass filter. Two objectives were used:

• 4x IR objective: single-tile FOV of 1.23×1.23 mm, 19.2 µm pixel size, mosaic of 7×6 tiles (8.6×7.4 mm), 4.3 GB data, 1 h acquisition
• 15x IR objective: single-tile FOV of 0.35×0.35 mm, 5.5 µm pixel size, mosaic of 27×19 tiles (8.4×6.7 mm), 42.4 GB data, 10.5 h acquisition

Data processing was performed in R with the hyperSpec toolbox. Pre-processing included removal of low-quality spectra, baseline correction, normalization and optional binning. A linear discriminant analysis (LDA) model was trained on 4x data and applied to both data sets for tissue classification.

Used Instrumentation

• Agilent Cary 620 FTIR microscope
• 64×64 Focal Plane Array detector
• Agilent Cary 670 FTIR spectrometer
• 4x and 15x infrared objectives

Main Results and Discussion

Both configurations provided comparable signal-to-noise ratios and uniform IR illumination in the 1200–1800 cm⁻¹ region. The 4x objective increased FOV by ≈3.4×3.4, reducing acquisition time by over 10× and data size by nearly 10×. LDA-based soft classification accurately discriminated five tissue classes (normal epithelium, connective tissue, inflammation, dysplasia, carcinoma/blood) in both data sets. The 4x resolution (19 µm) sufficed for margin detection and class mapping, while the 15x resolution (5.5 µm) was required to resolve fine structures (e.g., blood vessels) and access low-wavenumber bands (<1100 cm⁻¹) such as phosphate and carbohydrate vibrations.

Benefits and Practical Applications

• Rapid survey imaging over large areas with the 4x objective reduces time and data management burdens
• High-resolution imaging with the 15x objective enables detailed histological interpretation
• Flexible combinations support diverse workflows in biomedical research, QA/QC and industrial analytics

Future Trends and Opportunities

• Integration of larger FPA detectors (e.g., 128×128) to further expand FOV and accelerate acquisition
• Advanced under-sampling and binning strategies to optimize data volumes
• High-performance computing and cloud-based processing for large mosaic data sets
• Extension to reflective/transflection imaging for a wider range of sample substrates

Conclusion

The 4x IR objective offers an efficient approach for rapid, large-area FTIR imaging with adequate resolution for tissue classification, while the 15x objective provides high spatial detail and access to low-wavenumber biochemical markers. Together, they enable tailored workflows for comprehensive biomedical and analytical applications.

References

1. Beleites C., Popp J., Krafft C., Kansiz M. FTIR Microscopic Imaging of Large Samples with 4x and 15x Infrared Objectives: A Case Study of a Carcinoma Tissue Section. Agilent Technologies Application Note 5991-1363EN, 2014.
2. R Core Team. R: A Language and Environment for Statistical Computing. R Foundation for Statistical Computing, Vienna, Austria.