|
|
Line 11: |
Line 11: |
| {{font color|#FF1493|7.55 2.52 10.91 5.04 6.71 5.04 2.52 5.04 12.59 11.75 14.27 5.87 6.71 6.71 7.55 2.52 13.43 15.11 5.87 8.39 3.36 7.55 8.39 13.43 2.52 8.39}} | | {{font color|#FF1493|7.55 2.52 10.91 5.04 6.71 5.04 2.52 5.04 12.59 11.75 14.27 5.87 6.71 6.71 7.55 2.52 13.43 15.11 5.87 8.39 3.36 7.55 8.39 13.43 2.52 8.39}} |
| | | |
− |
| |
− | <!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd"><html xmlns="http://www.w3.org/1999/xhtml"><head><meta http-equiv="Content-Type" content="text/html; charset=utf-8" /><meta http-equiv="Content-Style-Type" content="text/css" /><meta name="generator" content="pandoc" /><title></title><style type="text/css">code{white-space: pre;}</style></head><body><h1 id="regression-testing-of-besa-mri-volume-conductor-segmentation">Regression testing of BESA MRI Volume Conductor Segmentation</h1>
| |
− | Regression testing for BM is done by comparing the volume conductor segmentation output between the old and the new version. Volume conductor segmentation is the last step in the segmentation pipeline of BESA MRI and any problems in the pipeline would be noticed when evaluating them. This comparison is done using Dice score. Dice Score is a standard metric used to measure the overlap between the 2 segmentation. The range of Dice score ranges from 0 to 1 indicating no overlap and complete overlap respectively. It is mathematically defined as <span class="math inline">$DSC = \frac{2\left| A \cap B \right|}{\left| A \right| + |B|}$</span>, where A and B are 2 matrices of equal size and |A| and |B| are the number of elements present in the respectively.<h2 id="testing-pipeline">Testing pipeline</h2>
| |
− | The testing pipeline reads the volume conductor segmentation data from the old and the new version and then compares them using Dice score.
| |
− | Data Required:<ul><li>
| |
− | Project data from Reference (old) version.</li><li>
| |
− | Project data from the version to be tested.</li><li>
| |
− | List of Project name and path to be tested.</li></ul><h3 id="pre-processing">Pre-Processing</h3><ul><li>
| |
− | Change the MATLAB code (BatchCompareFiles_MMYYYY.m) to reflect the correct data folder for Reference and Testing (DataFolderRef,DataFolderTest).</li><li>
| |
− | Create a 2 text files that contains the project name and its respective project path.</li><li>
| |
− | Make sure the files containing the list of project name and project path (ProjectNamesFilename,ProjectPathsFilename) reflect the actual location.</li></ul><h3 id="verification">Verification<img src="data:image/png;base64,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" width="239" height="192" /></h3>
| |
− | After completing the 2 steps mentioned above the MATLAB script for regression test can be executed. After the test is over, a Result.txt is generated that will have the result of the test. It will have Dice scores between all the pair of images for all tissue classes. Ideally all the 8 values should be 1.0.<h2 id="validation-of-test-script">Validation of Test Script</h2>
| |
− | To validate our test script, we verify the dice score of known 2D and 3D objects. Below is a list of all the cases that we have verified.<ul><li>
| |
− | <strong>2D object with single class</strong> – Here we generate a couple of 2D binary images (5x5). The Dice score is pre-calculated manually. These images then passed on to our pipeline and compared against the manually calculated value.</li><li>
| |
− | <strong>3D object with single class</strong> – Here we generate a couple of 3D binary object (10x10x10). The Dice score is pre-calculated manually. These objects are then passed on to our pipeline and compared against the manually calculated value.</li><li>
| |
− | <strong>3D object with multiple class</strong> – Here we generate a couple of 3D object (10x10x10) having 2 classes. The Dice score is pre-calculated manually. These objects are then passed on to our pipeline and compared against the manually calculated value.</li><li>
| |
− | <strong>Real 3D object with multiple class</strong> – Here we use 2 volume conductor segmentation with 8 classes. The Dice score is calculated using a known software package nipipe (<a href="http://nipype.readthedocs.io" class="uri">http://nipype.readthedocs.io</a>). These objects are then passed on to our pipeline and compared against the manually calculated value.</li></ul></body></html>
| |
| | | |
| | | |