- Wed Mar 04, 2020 8:32 am
The most important thing in the RT setup is to daily achieve the posture and the position of of the patient in the setup image to correspond the position and the posture of the patient in the reference image, as accurately as possible. In the breast case optimal situation is such that the entire bony structure on the chest wall (including vertebra and shoulder joint) and the breast tissue should locate typically inside +- 5 mm accuracy, depending on hospitals IGRT tolerances. At the time AlignRT has led the accuracy inside your given IGRT tolerances, you can rely on the system and skip the images or take only tangential images to just verify the accuracy. We sometimes have situations that we have to perform daily AP+LAT imaging protocol, since there is that much variation in the patient posture or isocenter. In FB breast cases this is not a problem with AlignRT. However, also in FB cases the goal is to go with the tangential images only, since it is 2-3 min. faster than AP+LAT and 4-5 min. faster than AP+LAT+tangential, which protocol is needed if there is displacement of the structures in the AP+LAT images and thereby difficulties to select the correct location for the online match and the couch shifts (which we finally want to lead to good accuracy in the tangential image). In addition, in some cases we have noticed that the system itself (with good setup and with good setup surfaces) can be more accurate than the rtts online match corrections based on AP+LAT images, to lead excellent accuracy in the tangential image ribs.
In my ideology, I prefer DICOM as a gold standard, just like your physicists. Still, rather often we notice that for reason or another we can not go on with the DICOM setup, since that much often we rtts notice it to lead small but annoying systematic errors, noticed during the first three fractions. The systematic errors can exist in the arm position, isocenter, vertebra rotation, displacement between breast and the bony structures, what ever. Based on the ROI comparison, the assumption that we have wrong kind of ROI and that is the reason why we can not stick with the DICOM seems not relevant. There is cases (not investigated how often) that we have to skip DICOM and acquire new surfaces for the setup. We have found that the arm position correction is easier and more accurate with the treatment capture than with the arm ROI. Which one do you use? In general I would like to setup the arm with the treatment capture where we see the entire arm up to the wrist, and also therefore I am not afraid of taking VRT surface for the setup. DICOM ends to mandibula level in CC direction
Now in your case you may need to find out is this arm position problem systematic or not, use offline image evaluation. If it is systematic, you could at the setup process with the treatment capture leave the arm to the posture you have found optimal (there is now offset in the treatment capture in the arm position or in the arm ROI delta towards the direction you evaluated from the images offline earlier) and if it was now ok in the images and everything else in the AP+LAT images was ok, acquire new gated reference capture for the next day setup. Verify with the tangential image that the soft tissue locates in its correct position also. Use this as a reference for the next day setup and acquire still at least once more AP+LAT (+tangential) image. Of course this procedure to acquire new reference surfaces and/or to verify the accuracy of new reference surface in the next day setup should in optimal cases be based on CBCT, where you can verify everything at the same time and even more reliable, but for example with the Varian linacs CBCTs you can not see both entire lymph node area and the soft tissue entirely with one scan at the images due to limited field size in CC direction, as far as I know...
So, do not consider DICOM as a gold standard for the setup, in the cases where it clearly leads to systematic errors in the images. RTTs are the one which daily setup the patients with the system and daily check the images online, so we are the best experts to say which part is working in those processes and which one is not. So bravely say those ills that you have noticed in the workflow in the meetings and discussions with the physicists, since finally the shared goal is the practical and accurate RT.
Thanks to this forum, we have been able to take along the workflow where we use one DICOM or VRT reference surface for the setup and go to the other field to acquire daily reference surface there for the treatment, no matter did we shift the couch based on online match or not. Hopefully you get other answers also and hopefully there was something useful. Have a nice day.
Best regards Marko