Related Experiment Video
Updated: Jul 17, 2026
11:05
High-definition Fourier Transform Infrared FT-IR Spectroscopic Imaging of Human Tissue Sections towards Improving Pathology
Published on: January 21, 2015
33.7K
Mid-Infrared Imaging Is Able to Characterize and Separate Cancer Cell Lines
E Kontsek1, A Pesti2, M Björnstedt3
12nd Department of Pathology, Semmelweis University, Budapest, Hungary. kontsek.endre@med.semmelweis-univ.hu.
Pathology Oncology Research : POR
|June 20, 2020
Summary
Label-free infrared spectroscopy effectively distinguishes cancer cells from normal cells. This dye-free method shows high accuracy, paving the way for improved intraoperative surgical guidance.
Area of Science:
- Biomedical Optics
- Spectroscopy
- Cancer Research
Background:
- Malignancies cause significant mortality, with macroscopic evaluation of surgical margins being unreliable.
- Emerging big data imaging techniques like mass spectrometry and various spectroscopy methods offer new diagnostic possibilities.
- While indocyanine green labeling is common, label-free mid-infrared imaging presents a more promising future for practical applications.
Purpose of the Study:
- To develop and evaluate a label-free infrared spectroscopy method for differentiating transformed (cancerous) and non-transformed (normal) cell lines.
- To assess the feasibility of using transflection spectroscopy in the mid-infrared range for cell classification.
- To identify optimal spectral regions and data processing techniques for accurate cell line separation.
Main Methods:
- Utilized a novel setup for label-free mid-infrared transflection spectroscopy on aluminum-coated glass slides.
- Analyzed both whole-range (4000-648 cm⁻¹) and fingerprint (1800-648 cm⁻¹) spectra of fixed cell lines (A-375, HT-29, CCD986SK).
- Applied mean transmission filtering, principal component analysis (PCA), and linear discriminant analysis (LDA) for spectral data processing and classification.
Main Results:
- Successfully excluded non-cell spectra based on mean transmission values (>90%).
- Demonstrated the feasibility of separating different cell lines using spectral filtering and chemometric analysis (PCA/LDA).
- Achieved high classification accuracy (99.84%) by focusing on the fingerprint spectral region within a specific mean transmittance range (70-75%).
Conclusions:
- The developed label-free infrared spectroscopy approach can accurately differentiate cell lines from various tissue origins in vitro.
- Effective data handling and spectral processing are crucial for adapting this dye-free technique for intraoperative use.
- Further research is essential to validate this marker-free approach in real-time surgical settings.
