3D ART VeSElecT

3D ART VeSElecT reconstructs vesicle structures from electron tomograms as a Fiji macro to enable automated detection, 3D segmentation, and morphological characterization of synaptic vesicles.

Key Features:

• Automated Vesicle Registration and Characterization: Integrates two macros within Fiji for registration and measurement of vesicles.
• Automatic and Semi-Automatic Modes: Provides both automatic and semi-automatic modes for vesicle detection and analysis.
• Manual Revision Step: Includes an optional manual revision step to minimize false positives and refine vesicle identification.
• Preprocessing and Contrast Enhancement: Processes image-stacks at 1 nm/pixel in x and y and applies contrast enhancement prior to analysis.
• Semi-Automated Cell Selection: Incorporates semi-automated cell selection to restrict analysis to areas of interest within tomograms.
• 3D Segmentation and Morphological Filtering: Performs 3D segmentation to detect vesicles and calculates vesicle volume and inner/outer diameters using a spherical approximation.
• Result Compilation with RoiManager Plugin: Organizes and color-displays results using the RoiManager plugin to allow identification and manual removal of confounding vesicles.
• Performance Evaluation Metrics: Assesses detected vesicle count, specificity (true vesicles), precision, and recall.
• Comparison with Learning-Based Methods: Demonstrates superior segmentation and morphological filtering compared to learning-based methods and reports up to 68× faster processing than manual annotation.

Scientific Applications:

• Developmental analysis in Danio rerio: Applied to analyze synaptic vesicles in embryonic Danio rerio at 4 and 8 days post-fertilization.
• Genetic mutation comparison in Caenorhabditis elegans: Used to compare wild-type neuromuscular junctions with the septin mutant unc-59(e261).
• Synaptic vesicle pool and morphology analysis: Identifies development-specific and mutation-specific changes in vesicle properties such as size and pool formation.

Methodology:

Preprocesses image-stacks (1 nm/pixel in x and y) with contrast enhancement; performs semi-automated cell selection; runs two Fiji macros to execute 3D segmentation for vesicle detection, calculates morphological parameters (volume and inner/outer diameters via spherical approximation), compiles results via RoiManager, and supports an optional manual revision step.