Attenuation-corrected fluorescence spectra unmixing for spectroscopy and microscopy

Area of Science:

• Biophotonics
• Spectroscopy
• Computational Imaging

Background:

• Fluorescence measurements are often distorted by light absorption and scattering.
• Conventional linear unmixing methods fail to correct for these attenuation effects.
• Accurate spectral separation is crucial for analyzing complex biological samples.

Purpose of the Study:

• To develop a new algorithm for fluorescence unmixing that accounts for attenuation-related spectral distortions.
• To improve the accuracy of separating overlapping fluorescence spectra in the presence of absorption and scattering.
• To extend the method for applications in spectral imaging microscopy.

Main Methods:

• A matrix-based approach was used to model the forward measurement formation.
• A corresponding inverse method based on nonnegative matrix factorization (NMF) was derived.
• The method was extended to a higher-dimensional tensor form for spectral imaging.

Main Results:

• The proposed algorithm successfully accounts for attenuation-related distortions in fluorescence spectra.
• Simulations and experimental results demonstrate superior performance compared to conventional linear unmixing.
• The method shows significant improvements when absorption and scattering are present, as in deep tissue imaging.

Conclusions:

• The developed fluorescence unmixing algorithm offers enhanced accuracy by correcting for light attenuation.
• This method provides a robust solution for spectral separation in complex optical environments.
• The tensor extension is valuable for advanced spectral imaging microscopy applications.