Enabling real-time reconstruction for large field-of-view single-molecule localization microscopy using discrete...

Area of Science:

• Biophysics
• Optical Microscopy
• Super-resolution Imaging

Background:

• Single-molecule localization microscopy (SMLM) achieves super-resolution by localizing individual fluorescent molecules.
• High numerical aperture objectives essential for SMLM exhibit distorted point-spread functions (PSFs) off-axis, limiting the field-of-view (FOV).
• These aberrations introduce significant 3D localization errors, restricting SMLM to small FOVs (e.g., 50 µm × 50 µm).

Purpose of the Study:

• To systematically investigate the impact of optical aberrations on large FOV 3D SMLM.
• To propose and validate a method for expanding the effective FOV in 3D SMLM.
• To enable real-time, high-precision molecular localization in large imaging areas.

Main Methods:

• Evaluated localization accuracy using unmodified, astigmatic, and double-helix PSFs under varying optical aberrations.
• Developed and tested discrete field-dependent PSFs for aberration correction.
• Employed GPU acceleration for high-speed image reconstruction of simulated and biological samples.

Main Results:

• Optical aberrations significantly degrade localization precision and accuracy at the edges of the FOV.
• Discrete field-dependent PSFs effectively correct for off-axis aberrations, enabling precise 3D localization across larger areas.
• GPU-accelerated discrete PSF models allow for real-time SMLM image reconstruction.

Conclusions:

• Optical aberrations are a critical limitation for large FOV 3D SMLM.
• Discrete field-dependent PSFs offer a viable strategy to overcome these limitations, enhancing FOV.
• Real-time reconstruction using GPU acceleration makes large FOV SMLM more practical for biological studies.