16 bit or 12 bit CT data
Peter Mohr
Surprisingly, the recent discussion on 16 bit or 12 bit CT data has disappeared in the online access to the topic "Siemens Med Tech Couch Model". Therefore, I have created this new topic, and below please find a repetition of the recent discussion.
Best regards,
Peter
From: pinnacl...@googlegroups.com [mailto:pinnacl...@googlegroups.com] On Behalf Of Mohr, Dr. Peter
Sent: Thursday, November 14, 2013 6:54 PM
To: 'pinnacl...@googlegroups.com'
Subject: [p3rtp] AW: 12 bit or 16 bit CT data
Dear Richard, dear all:
The answer the above question is clear (at least for me): 16 Bit is better than 12 Bit!
Pinnacle is able to work with 16-Bit data under several circumstances.
First, you have to extend your CT-Density-Table up to the requested densities. I use a maximum value of 30000 corresponding to density 16.0, and I have calibrated my table with a phantom with different insets including a thin titanium rod (in my CT Toshiba Aquilion LB: 8036 corresponds 4.59). In addition, I used a patient with known metal (in my CT: 15500 corresponds to 8.46). As soon as you have such an extended Density table, you can use higher densities than 4095 by the density override function (already without 16-Bit CT values).
Next, I have checked the transmission through 2 cm iron (a typical thickness for metal in patients), and Pinnacle exactly reproduced my measured transmission. So I am happy with Pinnacle for transmission through huge densities. Even if this is perhaps not perfect, it is definitely much better than using the truncated density 4095 ! (Maybe, there are some inaccuracies in the dose build-up at the border between normal density tissue and metal.)
Finally, I have found out that the 12-Bit truncation in Pinnacle only occurs during image import, i.e. in the conversion form the [CT-name].[slice].dcm files to the combined ImageSet_*.img file which is used in Pinnacle. I managed to write an import filter which does not truncate the CT data, but provides the full 16-Bit resolution. Then Pinnacle uses the 16-Bit CT data without major problems. (There are some minor problems when displaying DRRs, depending on the underlying DRR density table).
At our site we use the 16-Bit CT images for several years, and I feel much more confident than before when irradiating patients with metal in the body. Nevertheless, I still try to avoid or at least minimize radiation through metal to reduce uncertainties in the dose calculation.
Unfortunately, my hand-made import filter is not sufficiently well written to provide it over the mailing list, and it probably must be adjusted to the local DICOM file names of the CT slices. In addition, my import filter is not able to deal with all patient set-ups (standard HFS and FFS are OK). In case of interest, please contact me directly.
Best regards / Mit freundlichen Grüßen
Peter Mohr
Dr. Peter Mohr
Strahlenschutzbevollmächtigter
Leitender Medizinphysiker
Diakonie-Klinikum Schwäbisch Hall gGmbH
Diakoniestraße 10
D-74523 Schwäbisch Hall
Tel. 0791 753 4845
Fax 0791 753 4911
Email: Peter...@dasdiak.de
www.dasdiak.de| Richard Short |  Nov 14 |
Dear list servers
On our Philips Big Bore Brilliance CT scanner, we can reconstruct the CT data in either 12 or 16 bit format. From our initial testing, the data at CT is reconstructed and exported in 16 bit format, but when imported into Pinnacle version 9.0, it is displayed only in 12 bit format. Can anyone provide information or assistance on this issue?
Thanks
Richard Short