CLIJ-assistant

CLIJ-assistant enables GPU-accelerated construction and optimization of Image Data Flow Graphs for quantitative fluorescent microscopy image processing using the CLIJ2 library within Fiji.


Key Features:

  • GPU-Accelerated Processing: Executes image processing operations on graphics processing units via the CLIJ2 library to accelerate analysis of large image datasets.
  • Image Data Flow Graphs (IDFGs): Represents workflows as Image Data Flow Graphs that store operations, parameters, and connections.
  • Expert System Integration: Tracks the sequence of operations forming an image and suggests subsequent processing steps.
  • Dynamic Parameter Adjustment: Allows real-time adjustment of early processing step parameters with instantaneous visualization of effects on final results.
  • Undo Functionality: Enables rewinding parameter changes by retaining historical operations, parameters, and graph connections within the IDFG.
  • Batch Processing: Supports rapid batch processing of large image datasets for high-throughput analysis.
  • Reproducibility and Interoperability: Exports scripts in ImageJ Macro, Java, Jython, JavaScript, Groovy, Python, and C++ for use across ImageJ, Fiji, Icy, Matlab, QuPath, Jupyter Notebooks, and Napari.

Scientific Applications:

  • High-throughput fluorescent microscopy image analysis: Accelerates quantitative analysis of large fluorescent microscopy datasets using GPU processing and batch workflows.
  • Workflow optimization and parameter tuning: Uses IDFG-based real-time parameter adjustment and expert-system suggestions to refine image processing pipelines.
  • Reproducible cross-platform deployment: Produces language-specific scripts for deployment and reproducibility across ImageJ, Fiji, Icy, Matlab, QuPath, Jupyter Notebooks, and Napari.

Methodology:

Uses GPU acceleration via the CLIJ2 library within Fiji; represents workflows as Image Data Flow Graphs that store operations, parameters, and connections; incorporates an expert system that tracks operations and suggests subsequent steps; supports real-time parameter adjustments with immediate visualization of effects; exports scripts in ImageJ Macro, Java, Jython, JavaScript, Groovy, Python, and C++.

Topics

Details

Tool Type:
desktop application
Programming Languages:
Java
Added:
1/18/2021
Last Updated:
2/11/2021

Operations

Publications

Haase R, Jain A, Rigaud S, Vorkel D, Rajasekhar P, Suckert T, Lambert TJ, Nunez-Iglesias J, Poole DP, Tomancak P, Myers EW. Interactive design of GPU-accelerated Image Data Flow Graphs and cross-platform deployment using multi-lingual code generation. Unknown Journal. 2020. doi:10.1101/2020.11.19.386565.

Links