Download Database

Readme

PLUS_ScratchAssayAnalyzer

PLUS Scratch Assay Analyzer

The software comprises of a scratch assay segmenter written in MATLAB and analyzer tool which produces plots of the segmentation output, written in Python.
The software has been developed and tested with MATLAB R2019a and Python 3.12 under Debian GNU/Linux 11 (bullseye).

Reproduce the published results

Download

• Extracting the downladed software After extracting the downloaded software you see the following folder structure

  |-- plus_scratchassayanalyzer
       |-- experiments
       |-- scratchassayanalyzertool
       |-- scratchassaysegmenter

• Download the scratch assay image database: PLUS-ScratchAssay-DB.

• Download the supplemental materials. These contain the results from the external software tools. Put the extracted folder into your project folder:

  |-- plus_scratchassayanalyzer
       |-- experiments
       |-- scratchassayanalyzertool
       |-- scratchassaysegmenter
       |-- sup_results_ext_sw_tools

Prepare the experiments

• Go to the folder experiments.
• Edit the configuration and script files: Edit the path to the location of the PLUS-ScratchAssay-DB in the following files:
  • crop_data_images.sh:
    

  ROOT_DIR=../PLUS-ScratchAssay-DB/

  • experiment_rrpr.sh:
    

  IMG_DATABASE=../PLUS-ScratchAssay-DB/Raw
  GROUND_TRUTH_MASKS=../PLUS-ScratchAssay-DB/GroundTruthMasks

  • config_benchmark_analysis.yml:
    Add the path to the PLUS-ScratchAssay-DB for the ground truth masks at the parameter segmentation_results:

  gt:
    name: gt
    result_file: "../results/result_gt.txt"
    segmentation_results: "../PLUS-ScratchAssay-DB/GroundTruthMasks"

  Run the experiments

Benchmark experiment on PLUS-ScratchAssay-DB

• run the script experiment_rrpr.sh
  It uses the configuration files config_ef.txt and config_lg.txt and will produce a output folder results, where the result files and the resulting segmentation masks are stored for both configurations.

• run the script analyze_experiment_benchmark.sh
  It starts the python sofware analyzer and produces the benchmark plot.

Comparison for of external software tools

• run the script crop_data_images.sh
  It cropes the original raw images and ground truth masks and stores them in data_cropped in your project folder.

• run the script experiment_rrpr_cropped.sh
  It uses the configuration files config_ef_cropped.txt and config_lg_cropped.txt and will produce a output folder results, where the result files and the resulting segmentation masks are stored for both configurations.

• run the script analyze_experiment_benchmark_ext_sw_tool_comparison.sh
  It starts the python sofware analyzer and produces the performance comparison plot. It uses the result data of the external software tools from the sup_results_ext_sw_tools folder.

Contact

• Michael Linortner (Maintainer)