On 12.02.2013 10:47, Clarkson, Matt wrote:
> Hi there,
>
> I have a 2 directories full of DICOM slices from a GE CT scanner.
>
> In the first folder I have 342 slices, that I would expect to be 3 volumes of 
> 114 slices. However, the data is read in as 115 volumes. One of those 115 
> volumes has 114 slices. The remaining 114 volumes have 2 slices each.


> In the second folder I have 513 slices, that I would expect to be 3 volumes 
> of 117 slices. However, the data is read in as 2 volumes. Both have 117 
> slices, but 1 volume as 2 time points, and the other volume has 1 time point.
>
> My "expected" number of slices comes from the fact that the scan protocol had 
> 3 phases for contrast enhancement and if I ask micron to convert to nifti I 
> get 3 volumes… which seems OK.

Dear Matt,

unfortunately you ran into a missing detail of our DICOM reading..
I just documented the missing feature in
http://bugs.mitk.org/show_bug.cgi?id=14487

The short summary is: the reader needs to group files and does this by 
comparing UIDs/sizes/orientation/etc. In your case this is not enough to 
distinguish the three phases. From what I can see, "Acquisition Time" 
would be a good differentiator, but this cannot be generalized.

What happens next is that each group of files is separated into 3D slice 
stacks with equal slice spacing. Unfortunately one of your stacks is 
shifted a bit and creates the impression (for the reader) that a very 
small inter-slice distance is correct. However, it cannot find more than 
two slices and starts creating new slice stacks, each of the size of 2.

A proper fix would require us to first rethink how our GUI and reader 
work together, so it cannot be implemented as a quick bugfix. It does 
not help you much to know, but at least you have proved that there is 
demand for users to select proper interpretation.

My suggestion is to work around this issue by separating the files into 
different folders and load them individually.
OR, if this is a possibility for you, you could try to make your scanner 
output separate series for separate scans (with distinct UIDs).

Regards,
Daniel

-- 
Dr. Daniel Maleike                           Telefon: + 49 6221 647976 3
Mint Medical GmbH, Friedrich-Ebert-Straße 2, 69221 Dossenheim/Heidelberg
Geschäftsführer: Dr. Matthias Baumhauer
Registergericht Mannheim, HRB 709351

------------------------------------------------------------------------------
Free Next-Gen Firewall Hardware Offer
Buy your Sophos next-gen firewall before the end March 2013 
and get the hardware for free! Learn more.
http://p.sf.net/sfu/sophos-d2d-feb
_______________________________________________
mitk-users mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/mitk-users

Reply via email to