Probability Map Viewer: near real-time probability map generator of serial block electron microscopy collections

Area of Science:

• Microscopy and Imaging Science
• Computational Biology
• Bioinformatics

Background:

• 3D electron microscopy generates large datasets requiring efficient analysis.
• Semi-automated probability maps aid in identifying organelles and cellular features.
• Current workflows can be time-consuming, hindering rapid data interpretation.

Purpose of the Study:

• To develop a tool for expediting the review of probability maps from 3D electron microscopy data.
• To enable near real-time computation and visualization of segmentation results.
• To facilitate the optimization of image analysis strategies during data acquisition.

Main Methods:

• Developed Probability Map Viewer, a Java-based web application.
• The application allows selection and application of voxel classifiers on sub-regions of data.
• Results are visualized as overlays on raw data through a web browser.

Main Results:

• Probability Map Viewer provides near real-time feedback on segmentation.
• Users can interactively select classifiers and visualize results on various devices.
• The tool accelerates the image analysis workflow by enabling rapid experimentation.

Conclusions:

• Probability Map Viewer significantly speeds up the review of 3D electron microscopy data.
• It empowers researchers to optimize segmentation strategies interactively during imaging.
• This enhances the efficiency and informativeness of the image analysis pipeline.