Single-sample image-fusion upsampling of fluorescence lifetime images

Area of Science:

• Biophotonics
• Microscopy
• Computational Imaging

Background:

• Fluorescence lifetime imaging microscopy (FLIM) offers valuable insights into molecular interactions and biological processes.
• Current FLIM technology faces limitations in image resolution at high acquisition speeds due to detector and signal processing constraints.

Purpose of the Study:

• To introduce a novel computational approach for enhancing FLIM image resolution.
• To address the bottleneck of achieving high resolution in FLIM without compromising acquisition speed.

Main Methods:

• We present single-sample image-fusion upsampling, a computational super-resolution technique.
• This method integrates measurements from a low-resolution time-resolved detector and a high-resolution intensity-only camera from the same sample.
• Statistically informed priors encoding cross-modal correlations are used to solve the inverse retrieval problem.

Main Results:

• The developed method significantly enhances FLIM image resolution.
• It effectively bypasses the risk of hallucinated artifacts common in data-driven super-resolution techniques.
• The fused images show superior quality compared to standard interpolation methods.

Conclusions:

• Single-sample image-fusion upsampling offers a robust solution for improving FLIM resolution.
• This approach overcomes existing technological limitations in time-resolved imaging.
• The general methodology is applicable to other image super-resolution challenges involving multi-modal data.