Using splines for point spread function calibration at non-uniform depths in localization microscopy

Introduction

This repository contains the source code for the paper titled "Using splines for point spread function calibration at non-uniform depths in localization microscopy". The paper preprint can be found on Biorxiv. In this document, we describe the structure of the repository and how it can be used.

Repository setup

The repository consists of helper functions (in Python and MATLAB), diffspline code (Python) and the example notebook (Python).

• Helper functions contain all the necessary functions for data generation, processing, spline calibration and estimation. They include matlab data generation, service_functions, view_utils directories, and generatePSF.m file for data generation, test.m with calibrate.m files for spline calibration (see original code).
• The diffspline implementation is done in Python and can be found in dspline folder.
• An example of all the stages with the data generation, processing, spline calibration and estimation can be found in the example Jupyter notebook .

Dependencies

PSF localization process extensively employs fastPSF package. The installation procedure can be found on GitLab.

Moreover, the code generation and spline estimation code leverages MATLAB code, therefore, it is recommended to use MATLAB of the latest version. Since we call the MATLAB-implemented functions from Python interface, we require the installation of the MATLAB engine for Python.