Measurement-based estimation of global pupil functions in 3D localization microscopy

Area of Science:

• Optical microscopy
• Computational imaging
• Biophysics

Background:

• Accurate three-dimensional (3D) localization of molecules is crucial in super-resolution microscopy.
• Existing point spread function (PSF) models often fail to account for system aberrations, limiting localization precision.
• Experimental calibration of PSF is essential for reliable 3D localization microscopy.

Purpose of the Study:

• To develop and validate a phase retrieval procedure for experimentally calibrating the PSF in 3D localization microscopy.
• To improve the accuracy and precision of molecular localization by refining the PSF model.
• To provide a robust method for characterizing system aberrations and their impact on 3D localization.

Main Methods:

• Utilized maximum likelihood estimation (MLE) for phase retrieval to estimate a global pupil phase function.
• Applied the method to a microscope with a tetrapod PSF and a 6 µm axial range.
• Compared localization accuracy and precision against a model based on scalar diffraction theory.

Main Results:

• The phase-retrieved PSF model significantly improved the accuracy and precision of 3D localizations.
• Localization precision achieved with the new model approached the theoretical limits imposed by estimation theory.
• Demonstrated the reproducibility of the phase retrieval procedure for reliable experimental application.

Conclusions:

• The MLE-based phase retrieval procedure offers a superior method for PSF calibration in 3D localization microscopy.
• Experimentally calibrated PSF models lead to enhanced molecular localization performance.
• The developed method and provided code facilitate improved 3D super-resolution imaging.