Good evening Jose
Tangential image is the most important image in the breast RT. With that we verify the tangential fields accuracy, typically most importantly CLD (central lung distance) accuracy. We evaluate it with matching the image based on ribs. Our AL for the ribs is 4 mm in LAT/VRT and 5 mm in LNG. The purpose of this bony match is to check that lung is not radiated more than planned. On the other hand, about the BHL, image is not telling much if the vertebra is not visible. The other important thing is to check breast tissue residual error. It should locate with our AL inside 7 mm (VMAT) and 1.0-1.5 cm + always inside treatment field (FiF..). With orthogonal images we do not see breast reliably. With AlignRT we have found it practical in the setup to verify/setup the soft tissue before images, after the setup is done with the ROI on the rigid chest wall. In the tangential images heart border may be visible reliably or not, because sometimes it is difficult to say about the shadow in the image; is it rib or heart. We have 5 mm AL for the heart border in tangential image. These are the action limits we have with the tangential images for all of our breast patients. In the tangential images with RPM and with tattoo setup we only want to see the ribs, breast and heart, since AP+LAT images are taken daily because of large isocenter variation and thereby all the other bony areas of interest are visualized well in those images.
As mentioned earlier, with AlignRT DIBH we mostly nowadays acquire only daily kV-tangential images with large field size, since isocenter variation is that small that couch shifts in the online match are inside tolerances and we rely on our workflow with the AlignRT. In the images we have vertebra, shoulder joint, ribs, heart and most of the breast visible in those images. Additionally to former ALs: vertebra VRT 4mm, LNG 5 mm, shoulder joint 7 mm in n+. To give you one single result of our data from the direction we are discussing right now and about the case you ask of: in 3% of the fractions 3 mm was exceeded in the tangential image ribs in vertical direction after othogonal images and after possible couch shifts based on online match.
Then the question would I shift the couch based on tangential images. If we treat opposing tangential fields in FB whole breast (WB), yes: both LAT/VRT and LNG. We have never shifted AP/LAT direction in practice in that group, but that is the only group where we would have justification to do it. I do not recommend to do it anyway. If we treat opposing tangential fields in BH WB, I would shift only LNG based on those images. If we treat opposing tangential fields in FB/BH and additional fields for lymph nodes I would shift only LNG based on those images IF, note if I see also vertebra and shoulder joint in kV- images and those are inside tolerances given above after shift. The reason for these are that based on tangential image we do not know how much there is real isocenter error in VRT and LAT. There can be 5 mm error in the vertical isocenter and 5 mm error in the LAT isocenter (in the orthogonal images) and those compensates each others such that there is zero error in the tangential image ribs. So if we mostly correct vertical isocenter based on tangential images and error really is in LAT isocenter it is dangerous, when treatment is given from any other angle than those two opposing tangential fields. This can be harmful for the lymph node area for example, if in addition to LAT error on the background vertebra is rotated such that th1 and thereby medulla rotates towards lymph node treatment field, all poorly visualized in the tangential images. In DIBH opposing tangential field cases I think that if we have displacement in VRT/LAT direction, we can not be sure is the reason for error in isocenter or BHL and that is why I would not ever shift the couch in LAT/VRT in DIBH based on tangential images.
With AlignRT as you noticed we rarely shift the couch based on tangential images only in LNG. We usually acquire new BH surface for that day treatment, only to avoid struggling with the deltas, but do not use that surface on the next day setup. On the next day we instead acquire AP+LAT images with the daily used surface which we activated back to use in the record mode and may be based on those images acquire new reference surface for the upcoming setups, if it agrees with the yesterdays couch shifts and everything else is still ok in the images. I think, it is not reliable to use reference surface which was created based on tangential image.
Best regards Marko