This extension is very effective…hands up for this effort
I am about to do a scientific paper on this extension. I want to get the ethical approval for scientific research
1 Like
Thank you @mahmoud for the positive feedback. You can always direct mail me here on discourse, but as @lassoan said, it would be better to have an open discussion in one of the topics.
If you have questions or feature requests concerning the LungCTAnalyzer, please post them here. If you find bugs please open an issue here.
If you use the LungCTAnalyzer and 3D Slicer in a scientfic paper please cite us as mentioned at the bottom of the project page.
Thanks all …Of course, and from the ethics of research cited you and your project.
I want to get semi automated quantification yeild from table of Covid … especially covid Q , functional and affected percentage of lungs
But… I have some questions before I start quantification of the data for my scientific paper
The results are then superimposed to the CT 2D views in standard colors: “Bulla” = black, “Inflated” = blue, “Infiltrated” = yellow, “Collapsed” = pink and “Vessel” = red"
It is known that there are main radiological features in Covid CT images such as (consolidation , ground glass opacity and crazy pazing opacity)
I’m noted, these appearances give the pink color … which is translated is collapse.
Radiologically, the consolidation , ground glass opacity and crazy pazing opacity is considered an infiltration opacities
-
The concept of collapsed here…?? means that loss of aertion or atlactasis
Moreover, what’s the accurate mean of Infiltrated -
Another question this extension can be differentiated in consolidation, ground glass opacity and crazy pazing opacity in CT images.
Greetings …
Great, I am currently analyzing two local CT datasets of the first and second wave as well as the open source submillimetric COVID-19 CT dataset of @PaoloZaffino and the following HU thresholds worked for me:
Bulla / emphysema: -1050 > x <= -990
Inflated: - 990 > x <= -650
Infiltrated: -650 > x <= -400
Collapsed: -400 > x <= 0
Vessel: 0 >= x < 3000
“Infiltration” should include “ground glass opacity” and “crazy paving”.
“Infiltrated” includes “consolidation”.
I am not referring to the nonlinear “COVID-Q” any longer, it was suggested to better use “% affected lung” as a severity parameter.
GGO and crazy paving are difficult to differentiate yet because they share similar HU ranges. We would need to take their shape and size into account.
You could use SlicerRadiomics for this. I’d suggest doing a literature search since this has been well studied.
1 Like
There is modified HU reference … But the default reference is as
After, my investigation I think the default HU is more accurate than the above new modified referance values because the value as same as non-aeration alveolis ( refer to collapse) represent as consolidation
I wont to use default references to stars analysis the dataset …
Please, give me your opinions in these issue
I suggest to analyze a few of your patients with different thresholds, but in the end it may turn out that our predefined LCTA threshold may work well.
In many papers I see the authors using a visual score of the radiologist and comparing that score to the computed results.
1 Like
We are trying to break the routine and see the achievement of the extension in testing a field other than visual score of the radiologist
Hallo I encountered some obstacles, attached to the pictures
It seems to me that the extension is in the process of development…because I noticed that there is a difference in the results, for example:
I downloaded the 3D slicing software using the CT Lung Analyzer on two laptops
The first laptop gave me the COVID results as pictured 1111(image 1)
The second laptop gave me the COVID results as pictured 2222 (image 2)
Which of the results is correct?
whts the best parameter’s … % affected lung as a severity parameter or collapse or infiltrated ???
please helpe me
Best regards
Hello,
you are obviously using two different versions of the LungCTAnalyzer, probably in two different versions of Slicer.
The first result table you attached is the more recent one (with % affected), the second is from December 2020.
I recommend doing the analysis with the stable version of slicer (4.11). To get the latest distro of the LuncCtAnalyzer (LCTA) extension deinstall it , restart Slicer, install LCTA again, restart. The current program version of LCTA is 2.44. There is a version label at the top of the LCTA GUI.
Hi-
I am a graduate student hoping to use the software to analyze mouse tumor volumes. Anyone have any thoughts on if this would be possible? Having a hard time having the program correlate mouse anatomy with human. thoughts?
Hi,
The Lung CT Analyzer is focused on lung density measurements - what kind of tumors are you analyzing? How many tumors in how many species?
Rudolf
You can segment tumors using “Grow from seeds” effect as shown in this short tutorial:
Or, if the tumor is easily distinguishable from the surrounding tissues then you may can segment using less clicks, using “Local Threshold” effect:
1 Like
Hello,
I have been trying to segment a lung CT of a COVID 19 patient and following the steps in your videos I end up with a 3D segment of the lungs that is black and not colored as yours can you please advise?
Thank you in advance.
Which version of Slicer are you using?
Could you also check the version number of the Lung CT Analyzer?
Thank you for your reply.
The 3D slicer version I am using is 4.11.20210226.
The Lung CT Analyzer version is V 2.47.
And you are using the “Show/hide final segments in 3D” button to create 3D VIEW ?
Thank you for replying.
Before reaching this step, when I click on show/hide preview in 3D under the Thresholds section, the entire 3D segment is black. I can change the opacity but the 3D model remains mono-colored. When I select the show/hide final segments in 3D under the Statistics section, the output colored regions are displayed but since the model was originally black the combination appears non-homogenous and distorted somehow.
