Page 1 of 1

CBCT DICOM REF IMAGE vs TX DICOM REF IMAGE

Posted: Wed Nov 07, 2018 7:12 pm
by sheila_cioffa
When treating Brain cases (SRT in our center), if the lesion is located far from mid-brain, we center the patient prior to the CBCT so as to acquire the full cranium.  When using VisionRT, you want to adjust based on your Treatment Dicom Reference surface.  What is your workflow in such a case as this?  Do you send over 2 DICOM Reference Surfaces, one for imaging and another for treatment?

Re: CBCT DICOM REF IMAGE vs TX DICOM REF IMAGE

Posted: Mon Mar 04, 2019 7:20 pm
by michael_tallhamer_msc_dabr
We use this workflow all the time for SRS/SRT/SBRT and other treatment types that use CBCT imaging with a couch centering in order to monitor the patient during imaging to make sure they do not move before, during, or after imaging and before applying the shift.

When couch centering is anticipated during imaging for an SRS/SRT/SBRT type of case (or any type of case for that matter) we create a "centering" field during the patient import in the Record work space.

You can either re-import the DICOM structure set into that field and re-import the ROI that you had drawn on the initial imported surface (i.e. from the ISOCENTER FIELD) or leave the field blank with no surface and ROI. If you choose, as we do, not to import the DICOM surface again you will have to draw an ROI on the fly in the treatment work space before monitoring the patient for motion during CBCT imaging.

During SRS treatment we will setup to the DICOM surface in the setup field (i.e. ISOCENTER FIELD) then once we are at zero offset from the DICOM (or really really close) we take a reference image to zero out at the setup position. We then start the CBCT procedure which will require the couch centering. Center the couch and switch to the centering field you created during import from above. At that point we take a reference image under the "centering" field and draw a quick ROI to monitor for patient motion before during and after imaging.

Once the CBCT is complete and the shifts are determined through fusion you can restore the couch (switching back to you previous reference image at the setup position) and then apply the shifts. Once the shifts are applied you can just take an additional reference image at the final position and use it for monitoring throughout the treatment. Very quick and easy

Re: CBCT DICOM REF IMAGE vs TX DICOM REF IMAGE

Posted: Tue Mar 05, 2019 7:48 am
by Annika Sartz
That was a useful and detailed description of the center couch CBCT workflow. When  performing a CBCT like this, do you ever turn into camera blockage by gantry or kV equipment? Do you sometimes treat your SBRT cases in DIBH? If you do DIBH, how do you handle a potential camera blockage in a DIBH/CBCT case? Do you define your CBCT-ROI according to a 'dry run' before treatment or any other workaround?

Re: CBCT DICOM REF IMAGE vs TX DICOM REF IMAGE

Posted: Wed Mar 06, 2019 2:44 pm
by michael_tallhamer_msc_dabr
As for the blockage question...

We perform an obstruction test every morning to verify the obstruction with kV arms extended (Varian TrueBeam) during CBCT does not impact the deltas during imaging for an ROI of the same size and location of an intracranial SRS case.

For SBRT it can be a bit more complicated based on your selection of the site type (i.e. Chest vs SBRT Lung for example) since that will impact the surface resolution and also the choice of the ROI size and/or location relative to isocenter. By that I mean if the surface is high resolution and the ROI used for monitoring during CBCT acquisition is large you will see more obstruction issues during CBCT primarily at head positions blocking the camera pods since the imaging panel will essentially be blocking the other lateral pod at the same time. With ROIs similar in size and relative isocenter location to intracranial SRS cases this isn't (or shouldn't be) much of an issue if your system is working properly. However with very large ROIs and relative isocenter positions that put the ROI "deep" (i.e. in toward the gantry relative to the isocenter) you can obstruct a higher percentage of the ROI by area during CBCT acquisition which will result in obstruction issues with the deltas.

To avoid this for SBRTs in the abdomen and chest you can do a couple things.

Use a lower resolution surface (i.e use Chest or Breast instead of SBRT Lung) for much larger ROIs. This may cause "jumpiness" in the deltas when the obstruction ratio of the ROI area exceeds the threshold given all other variables are held constant but the higher refresh rates will bring it back much quicker when the ROI is less obstructed. I don't typically address it completely with this method but some have very strong feelings about ROI size and location vs surface resolution vs accuracy so I include it as a standalone option here
Create / use an ROI that is for CBCT acquisition monitoring "ONLY" that moves its relative position to isocenter to a more favorable or shallow position (i.e. away from isocenter toward the foot of couch). This can mean moving an entire ROI down on the patient for CBCT monitoring or changing the shape of an ROI (still over isocenter) but with more surface area toward the foot of the couch away from isocenter so you compensate for the percentage of the ROI seen during CBCT outside of the panel/head obstruction zones.

This is my preferred method.
This sometimes means you are creating an ROI over an area of the patient that is "less ideal" for monitoring for say treatment due to breathing motion.
Remember ROIs are "free" and you can have as many as you like so create a few spares to play with in the treatment screen and use one for imaging monitoring and another more ideal for treatment monitoring.
You can also use both...let that sink in a minute...by that I mean if you feel the treatment ROI is preferred for monitoring say because motion over that region is independent of a region more visible during CBCT and you are worried it is going to move independently during imaging...you can monitor with one and when you see obstruction switch to the other more visible ROI "on the fly" just during that portion of the CBCT and then back after the obstruction zone is passed. That should be very rare but I have see a few cases in 6 or so years where that could be used.
Also, remember you have gated captures to help compensate for breathing motion if it is an issue during imaging or treatment so the solution may require a multifaceted approach.


Some combination of the above 2 solutions is often the ideal approach but is not without it's trade-offs in ROI size/location or resolution/accuracy so weigh your options in light of other imaging that may be used during treatment (like triggered imaging) and the true purpose of the surface monitoring during the treatment at hand.

As for DIBH SBRT...

We use the same procedure for determining ROI adjustments for couch centering (If I understand that portion of the question correctly) but we obviously have to modify the CBCT couch centering procedure to capture a DIBH reference at the couch centered position because it only exists at isocenter in the DICOM reference. I believe I have outline that process on the forums in the past but can do it again here on this thread if that is preferred later.

Finally CBCT-ROI definition...

In current versions of the software you don't have to do anything special if you only want a single ROI to work with at the couch centered position. Just create the centering field in the record work space and the capture a reference image either on day 1 of treatment or on the vsim day. Once the centered reference is captured under the centering field you can draw the Default ROI in the treatment workspace.

If you want to have the option to play with a couple ROIs at the couch centered position you can import the DICOM surface into the centering field you create in the record work space and add as many as you like there and "pre-draw" them so they auto propagate when taking the reference on the first day so that there will be minimal work at the machine needed to edit/create these "on the fly" so to speak. Either way is acceptable you just have to determine what fits better into your workflow.