Locally adaptive thresholding centroid localization in confocal microscopy.

Area of Science:

• Optics and Photonics
• Biophysical Techniques
• Microscopy

Background:

• Confocal microscopy requires precise localization of axial response signals for accurate measurements.
• Existing methods, such as maximum likelihood estimation, can be computationally intensive and slow, especially with sparse data.

Purpose of the Study:

• To develop a novel, fast, and accurate algorithm for nanometer-level peak position localization in confocal microscopy.
• To achieve localization accuracies near theoretical limits, particularly for signals with few sampling points.

Main Methods:

• An iteration-free centroid algorithm incorporating a locally adaptive threshold was developed.
• The algorithm's performance was evaluated using both simulations and experimental data.

Main Results:

• The proposed algorithm demonstrated nanometer-level localization accuracy for peak positions.
• Accuracy approached theoretical limits, especially when analyzing discrete signals with a small number of sampling points.
• The algorithm was significantly faster (orders of magnitude) than traditional maximum likelihood estimation fitting schemes.

Conclusions:

• The developed centroid algorithm provides a highly accurate and computationally efficient solution for axial response signal localization in confocal microscopy.
• This method is particularly advantageous for applications with limited sampling points, offering near-optimal localization performance.