Illuminating native fluorescence signatures for cancer detection

Area of Science:

• Biomedical Optics
• Cancer Research
• Molecular Imaging

Background:

• Autofluorescence, light emission from biological substances, changes with cancer-related tissue alterations.
• Endogenous fluorophores like collagen, NADH, FAD, and amino acids exhibit spectral shifts during colorectal cancer (CRC) progression.
• Current hyperspectral imaging methods (Em-HSI) are too slow for real-time endoscopic screening.

Purpose of the Study:

• To quantitatively characterize endogenous fluorophores in mouse tissues.
• To establish a reference spectral library for spectral unmixing of hyperspectral images.
• To develop a faster hyperspectral imaging approach for CRC detection.

Main Methods:

• Spectrofluorometry and fluorescence microscopy were used to characterize endogenous fluorophores.
• Excitation-scanning hyperspectral imaging (Ex-HSI) was employed for spectral data acquisition.
• A spectral library of biologically relevant autofluorescent molecules was created.

Main Results:

• Comprehensive spectral data for key endogenous fluorophores (collagen, elastin, NADH, FAD, PPIX, tryptophan, tyrosine) were obtained.
• The study successfully established a reference spectral library for mouse colorectal cancer tissues.
• The Ex-HSI approach demonstrated potential for faster hyperspectral image acquisition.

Conclusions:

• Autofluorescence properties are significantly altered by colorectal cancer progression and tumor microenvironment changes.
• The developed spectral library and Ex-HSI method can enable spectral unmixing for improved CRC diagnosis.
• This work paves the way for real-time, non-invasive endoscopic screening for colorectal cancer.