Model-free machine learning-based 3D single molecule localisation microscopy.

Area of Science:

• Biophysics
• Microscopy
• Computational Biology

Background:

• Single molecule localization microscopy (SMLM) offers super-resolution imaging but 3D SMLM typically requires microscope modifications.
• Existing 3D SMLM methods often demand significant computational resources and time.

Purpose of the Study:

• To present an accessible and efficient 3D SMLM method applicable to standard fluorescence microscopes.
• To develop a computationally lightweight approach for high-throughput 3D super-resolution imaging.

Main Methods:

• Development of 'easyZloc', a 3D SMLM technique utilizing a lightweight Convolutional Neural Network.
• Application of the easyZloc method to standard, unmodified fluorescence microscopes.
• Demonstration of 3D reconstruction capabilities for biological samples.

Main Results:

• Achieved comparable performance to existing methods in reconstructing nuclear pore complexes.
• Demonstrated significant reductions in computational power and execution time.
• Successfully performed 3D reconstructions of the nuclear envelope and actin samples over extended axial ranges.

Conclusions:

• The easyZloc method provides a practical and efficient 3D SMLM solution for standard microscopes.
• This approach is suitable for high-throughput workflows, reducing computational barriers in super-resolution microscopy.
• easyZloc expands the accessibility of 3D super-resolution imaging in biological research.