Template-free 2D particle fusion in localization microscopy.

Area of Science:

• Biophysics
• Microscopy
• Nanotechnology

Background:

• Super-resolution microscopy techniques like localization microscopy offer high resolution but can be limited by template bias and registration errors.
• Image fusion methods aim to improve signal-to-noise ratio and resolution but often inherit limitations from individual image registrations.

Purpose of the Study:

• To develop a robust, template-free particle-fusion approach for super-resolution microscopy.
• To overcome the challenges of template bias and registration sensitivity in image fusion.

Main Methods:

• A novel all-to-all registration algorithm was developed for particle fusion.
• The method was applied to DNA origami nanostructures with varying labeling densities.

Main Results:

• Achieved 3.3-nm Fourier ring correlation (FRC) resolution by fusing 383 DNA origami nanostructures with 80% labeling density.
• Attained 5.0-nm resolution for structures with only 30% labeling density, demonstrating robustness to underlabeling.

Conclusions:

• The template-free particle-fusion approach enhances resolution and signal-to-noise ratio in localization microscopy.
• This method offers improved robustness against registration errors and underlabeling, broadening its applicability.