Open berolling opened 6 years ago
Could you share the three components of the data,which is named CT,Strain and AirwayDTF,I want to refer it. In addition, I want to know how these three kinds of data are integrated into data,I would appreciate it if you could help resolve this problem .
@lsoaringl hi, CT is just the raw data of CT scan; Strain is a vesselness feature extracted using strain energy filter based on the CT scan, (the filter is also open source); AirwayDTF is a feature which described the distance to airway, where the airway is extracted using region growing methond in CT scan.
Thank you very much for replying to me. I also want to know how these three kinds of data are combined into “data.mat”. I want to reproduce this experiment using my data.
@lsoaringl hi, If you got three data: CT, Strain and AirwayDTF, it is not necessary to make it into a structure data.mat.
You can replace this part:
Data = load('data.mat'); CT = im2ones(Data.CT); % CT intensity Strain = im2ones( Data.Strain ); %vesselness from strain energy filter: pxenhancement of https://github.com/ITKTools/ITKTools AirwayDTF = Data.AirwayDTF;
with
CT = im2ones(read_CT_file); % CT intensity Strain = im2ones(read_Strain_file ); %vesselness from strain energy filter: pxenhancement of https://github.com/ITKTools/ITKTools AirwayDTF = read_AirwayDTF_file;
Hello Sir... Can you please tell me where the open source code is available of strain energy filter. I couldn't find it... I did install mevislab for Airway DTF using region growing but first I have to extract vessel from the vessel12 dataset. If you please share link of the source code. I do read paper Xiao, C. et al. Medical Image analysis, 15(1):112-24. 2010.
@SMKhan11 Hi, the details of implementation can be found in the paper, 'Automatic quantitative analysis of pulmonary vascular morphology in CT images'. The source code of 'strain energy filter' can be found in https://github.com/ITKTools/ITKTools.
@chushan89 sir you have used visual studio IDE for strain energy filter with cmake build?
@SMKhan11 There is a README in ITKTools, which you can follow. I used visual studio 2010 or 2013 for building.
Hi Sir. I have got the strain energy filter based on your comments. Could you please tell me the details of the AirwayDTF? I have installed MeVislab but I am not sure whether I should use RegionGrowingMacro or DtfSkeletonization. Thank you very much.
@xuanxu92 With RegionGrowingMacro, you need to identify a seed of bronchi tree in CT, and the lower & upper threshold should also be adjusted. If the settings are adjusted properly, a coarse bronchi tree could be detected. Then, the 'EuclideanDistanceTransform' is used to generate the AirwayDTF feature.
@xuanxu92 how you got the stain energy filter, i am stuck in this step bcz of visual studio and i dont know how to import these external libraries https://github.com/ITKTools/ITKTools. in ITK tool visual studio. plz help
@xuanxu92 how you got the stain energy filter, i am stuck in this step bcz of visual studio and i dont know how to import these external libraries https://github.com/ITKTools/ITKTools. in ITK tool visual studio. plz help
hi, I use MacBook and python, so I don't import the ITK tools into library. I install the ITK first, then ITK tools.
@Ingrid1989 If the ITKtookit were downloaded and were successfully build & compiled, the executable file of strain energy filter can be called in cmd or python.
Then, the 'EuclideanDistanceTransform' is used to generate the AirwayDTF feature.
I Have created a mask for the coarse bronchi tree with mevislab. However, when I use the 'EuclideanDistanceTransform' It looks really strange (maybe I need some parameters to be changed). My question is, should I apply the bronchi tree mask as a overlay of the original CT or do I extract them? And which parameters from the 'EuclideanDistanceTransform' should be changed? And do you already mask the image with a lung mask before determining of the EuclideanDistance?
@Ingrid1989 Hi, The 'RegionGrowingMacro' of Mevislab was used to extract a coarse bronchi tree. Based on the extracted airways, 'EuclideanDistanceTransform' module was used to generated the distance map to airway, where dimension was set to 2d (you could also try with 3D); define foreground with min value=1 and max value=1; the default settings were used for rest parameters. Best, Zhiwei
Does Anyone else have problems with making the ITK toolbox run in visual studio 2019? It keeps saying it can't find the CppCodeProvider.dll no mather what I try...
@Ingrid1989 If there are problems with using ITKtools, refering to https://github.com/ITKTools/ITKTools may get useful answers. Good luck with your problem. Best, Zhiwei
hi, I use MacBook and python, so I don't import the ITK tools into library. I install the ITK first, then ITK tools.
How did you install them? Because ITK is recognised by pip install itk, but itktools is not.
As I known, the itktools can't be installed using pip install. If it was builded and compiled, then you can use these functions by calling the excutable.
Queation: I want to ask the data in the source code which is named Data. It is a struct combined by CT, Strain and AirwayDTF. Could you tell me how can I get the Strain and AirwayDTF of the new img which is 3D CT images? The answer of author: The ‘Strain’ features was extracted using strain energy filer (Xiao, C. et al. Medical Image analysis, 15(1):112-24. 2010). The AirwayDTF is a feature which described the distance to airway, where the airway is extracted using region growing methond in Mevislab.