[IGSTK-Developers] Re: Multi-Frame DICOM

[Mathieu Malaterre]

Luis Ibanez wrote:
> 
> Hi Mathieu,
> 
> We are wondering on what could be the appropriate
> way to find whether a DICOM file is a multi-frame.
> 
> We are facing this in an application in IGSTK.
> 
> Could you please give us some hints ?

What is being used in GDCM is the following (*). So you need to parse 
the header, and retrieve the tag: 0028|0008. If found AND different from 
1, then yes it should be a multi frame images.
For more info, please refer to PS 3. C.7.6.6. Multi-Frame Module.

But as David Gobbi recently stated some DICOM can be 4D (Nuclear 
Medicine Modality) in a single file, in which case you need to verify 
the Value Multiplicity of "Frame Increment Pointer" (0028|0009).
This might cause trouble as it is not supported by GDCM.

HTH
Mathieu

(*)
int File::GetZSize()
{
    // Both  DicomV3 and ACR/Nema consider the "Number of Frames"
    // as the third dimension.
    DataEntry *entry = GetDataEntry(0x0028,0x0008);
    if( entry )
       return (int)entry->GetValue(0);

    // We then consider the "Planes" entry as the third dimension
    entry = GetDataEntry(0x0028,0x0012);
    if( entry )
       return (int)entry->GetValue(0);
    return 1;
}