Using pydicom in MatLab

I'm experimenting with using pydicom functions inside MatLab as an alternative to dicominfo to speed up reading of dicoms.

I import the module and read the dicom header of the file dcmFile using dcmread:

py.importlib.import_module('pydicom');
ds=py.pydicom.dcmread([dcmFile],false,true);

It works really well, but how do I efficiently extract single tags from the ds object?

As an example, in python, I would use ds[0008,103e] or ds.SeriesDescription to get the Series Description tag. But this doesn't work in Matlab for the Python FileDataset object.

The following works, but is rather cumbersome:

tmp=ds.data_element('SeriesDescription'); seriesDescription=char(tmp.value);

Is there a better way to extract tags?

Thanks,

Lars

[Rik]

You could consider using parts of the Dicom Toolbox that Dirk-Jan Kroon published on the FEX. I am also working on that, but my code is not yet in a shape where I'm comfortable publishing it on the FEX myself. The code by Dirk-Jan is less robust, but it often at least an order of magnitude faster than dicomread and dicominfo.

[Lars Tolbod]

Thanks, Rik. But I'm not sure I understand why it should be faster. As far as I can tell the Dicom Toolbox is also just using dicominfo from the Image Processing Toolbox?

[Rik]

I assumed he would be using his same code everywhere, apparently not. This submission doesn't use the builtin tools. You might want to use the code below to create the dictionary.

%%replace this
% Load Dicom Tag Library
functionname='ReadDicomElementList.m';
functiondir=which(functionname);
functiondir=functiondir(1:end-length(functionname));
load([functiondir 'Dictonary/DicomTagDictionary.mat']); 

%%with this
dict_base=ingest_dict;
dcmdic.group=zeros(size(dict_base,1),1);
dcmdic.element=zeros(size(dict_base,1),1);
for n=1:size(dict_base,1)
    dcmdic.group(n)=hex2dec(dict_base{n,1});
    dcmdic.element(n)=hex2dec(dict_base{n,2});
end
dcmdic.type=dict_base(:,3);
dcmdic.name=dict_base(:,4);