A modular framework for multi-scale tissue imaging and neuronal segmentation.

Area of Science:

• Neuroscience
• Biomedical Imaging
• Computational Biology

Background:

• High-resolution 3D imaging of neurons generates vast datasets, but tools for analyzing cellular and sub-cellular structures are lacking.
• Challenges include high neuronal density, low signal-to-noise ratios in thick samples, and data heterogeneity from various imaging methods.

Purpose of the Study:

• To develop a robust framework for high-resolution imaging and analysis of neuronal structures.
• To create a scalable algorithm, SENPAI, for segmenting neurons at cellular and sub-cellular scales.

Main Methods:

• Established a sample preparation method for high-resolution imaging of thick brain tissue.
• Developed SENPAI, a scalable algorithm for segmenting neurons in conventional and super-resolution STimulated Emission Depletion (STED) microscopy images.
• Proposed a novel validation paradigm for segmentation performance assessment.

Main Results:

• SENPAI achieves accurate multi-scale segmentation, encompassing entire neurons and individual spines.
• The algorithm demonstrates superior performance compared to existing state-of-the-art segmentation tools.
• The developed framework enhances the processing of complex neuronal circuitries.

Conclusions:

• The SENPAI algorithm and associated framework provide a powerful solution for neuronal structure segmentation.
• This advancement will significantly aid researchers in analyzing intricate neuronal circuitries from high-resolution microscopy data.