Google Groups no longer supports new Usenet posts or subscriptions. Historical content remains viewable.
Dismiss
Toshiba MR Private Tag (0029,1090)
380 views
Skip to first unread message
Chad
Feb 12, 2015, 5:05:16 PM2/12/15
to
I've got some diffusion tensor data from a Toshiba Titan 3T scanner (software V2.30*R005), but I did not have the gradient orientations. They did not provide a (0018,9089) Diffusion Gradient Orientation tag.
They have a private SQ (0029,1001) that holds 6 items each with an OB attribute (0029,1090). After looking at the hex dump, I determined that there was more DICOM-ish data in it. I had to reverse the bytes and swap nybble-pairs, but after doing so, I was able to read each (0029,1090) attribute as a DICOM object. The 5th item held a few useful non-private tags including my needed (0018,9089) - and the DTI's look nice.
I've been unable to confirm from Toshiba that this is consistent between scanners/software, but was hoping someone here could. I realize this is private tag data and all bets are off, but I'd feel more comfortable if there was another data point besides this one scanner.
My java code to do the data adjustment is:
private static byte[] adjustToshibaBytes( byte[] input )
{
int len = input.length;
byte[] output = new byte[ len ];
for( int i = 0; i < len; i++ )
{
output[ i ] = (byte)((input[ len - i - 1 ] & 0x0F) + (input[ len - i - 1 ] & 0xF0));
}
return output;
}
Hope this is useful to someone else as well.
-Chad
They have a private SQ (0029,1001) that holds 6 items each with an OB attribute (0029,1090). After looking at the hex dump, I determined that there was more DICOM-ish data in it. I had to reverse the bytes and swap nybble-pairs, but after doing so, I was able to read each (0029,1090) attribute as a DICOM object. The 5th item held a few useful non-private tags including my needed (0018,9089) - and the DTI's look nice.
I've been unable to confirm from Toshiba that this is consistent between scanners/software, but was hoping someone here could. I realize this is private tag data and all bets are off, but I'd feel more comfortable if there was another data point besides this one scanner.
My java code to do the data adjustment is:
private static byte[] adjustToshibaBytes( byte[] input )
{
int len = input.length;
byte[] output = new byte[ len ];
for( int i = 0; i < len; i++ )
{
output[ i ] = (byte)((input[ len - i - 1 ] & 0x0F) + (input[ len - i - 1 ] & 0xF0));
}
return output;
}
Hope this is useful to someone else as well.
-Chad
Mathieu Malaterre
Feb 24, 2015, 12:12:38 PM2/24/15
to
https://sourceforge.net/p/gdcm/gdcm/ci/e0b13cf77358692501aaacb0c31853c8b6adb436/
I simply needed to read the binary blob in the opposite order. Then
the dataset is strictly an explicit little endian one. Thanks for the
datasets.
And I was able to deduce some undocumented TOSHIBA attributes:
http://sourceforge.net/p/gdcm/gdcm/ci/9ff644cf464a8f752e32f8b29063a4732b822a1b/
with an issue:
http://sourceforge.net/p/gdcm/gdcm/ci/e21590256ffc45c036f9db15e54ba292b41bf25b/
Looks like with the dataset chad sent me there is an inversion in the private creator which leads to duplicate attribute (0019,002a,PMTF INFORMATION DATA^12)
but no (0019,002a,PMTF INFORMATION DATA^10), which I suspect is the cause of duplication.
0 new messages