LungCTAnalyzer error: Input lung CT is invalid

Hi @ashipde,
it would be great if you could do another test with an update to Lung CT Analyzer I pushed to Github today. It should be automatically available as an update in the extension manager tomorrow or the day after and should hopefully fix the issue with LTCA.ini you experienced. No hurry, just to be sure that the permission error has been solved. It should show V 2.50

Concerning your ILD study I will be glad to help, even may be able to provide a few ILD cases if you are interested.

Thank you

@rbumm Dr. Bumm, the updated extension on GitHub works well. The version number for Analyzer also gets displayed now (as 2.49). I will test the in-app extension once it gets available (new version not yet there, at 8 pm UTC).

I thank you and Dr. Lasso for the suggestions and offer of help for the ILD analysis, and I may request advice/help.

@rbumm Dr. Bumm, two questions:

(1) Can LungCTAnalyzer be used on multiple CT series of same subject sequentially, or does one have to delete some nodes in between different CTs, or restart the app altogether? I am asking because, in my system, some nodes like results and masked volume that are created after examination of a series get put directly under subject and not the CT series in the node hierarchy.

(2) Is rescaling of any sort is needed before using LungCTAnalyzer, and if I were to leave default module thresholds, will my results be to inaccurate or just off by a minor fraction?

(1) You can analyze a series of CT data (f.e. within a directory) and save the result to a CSV file line by line automatically. However, you would need to segment the lung masks for each volume beforehand. Please have a look at SlicerLungCTAnalyzer/PythonScripts at master · rbumm/SlicerLungCTAnalyzer · GitHub.

(2) this depends on your data. If it is all from one CT machine I would leave the thresholds as they are.

@rbumm Thank you Dr. Bumm for the suggestions. Regarding thresholding, is it then an ‘art’ that one learns by experience, or are there some sound principles that one can use? I am asking this because I am trying to use multiple CTs from different machines to attempt to track ILD progression. Thank you.

Not exactly an “art” … I would suggest defining the range of the sliders in Lung CT Analyzer in a way that the preview covers the interstitial abnormalities as best as possible with the ranges of “Infiltrated” and “Collapsed”. Find a compromise here, and maybe discuss it with your radiologist. Then run the analysis with unchanged sliders over all datasets and evaluate the difference between the initial and follow-up data (% change) for each individual patient. Would be really cool if you could quantify a progression that way.
Hope that helps …

1 Like

@rbumm Dr. Bumm, will contrast affect LungCTAnalyzer analysis. If so, which aspect - infiltrate, consolidation, etc. - is likely to get exaggerated? Please also comment on the effect of depth of inspiration.