Some time ago here was a discussion on the forum about diaphragm level displacement in DIBH. I think these results fits well under this topic also. After that question I did IGRT study, where I compared interfractional displacement 1) between the vertebra th6 and the sternum in VRT and LNG and displacement 2) between vertebra th 6 and the diaphragm level in LNG, all evaluated from the LAT kV-images. Patients (n=25) were treated with RPM quidance, marker box placed on the sternum with window level +-2mm. Displacement between vertebra and sternum (BHL) in the group were (Σ, 0.22 cm, σ 0.16 cm in VRT and Σ 0.37 cm, σ 0.26 cm in LNG) demonstrating typical BHL values in DIBH studies with RPM. Displacement between vertebra and diaphragm in that group in LNG was (Σ 0.64cm, σ 0.42 cm). Errors between vertebra vs. diaphragm were typically much larger than the errors between vertebra vs. sternum in LNG. There were even 2 cm residual displacement error in the diaphragm level, even though the BHL (vertebra vs. sternum VRT and LNG) was ok. This means that what happens on the bony chest wall in DIBH does not necessarily correlate with the internal anatomy interfractionally. The theoretical reasons are related to diaphragms anatomical day to day variations and differences in the breath holds which are not necessarily visible in the bony structures. This knowledge has not much to do with the variation of the diaphragm compared to sternum/chest wall between the BHs intrafractionally, since this was not investigated.
Some words about breast DIBHs concerning belly breathing. With breast DIBHs we use RPM at the CT, since there is five other linacs in the hospital where AlignRT is not installed and with those all DIBHs are still treated. At the CT, RPM box is placed on the rigid sternum, since it is investigated and published that such box location leads to improved accuracy in the realized BHL evaluated from LAT images (vertebra vs. sternum). Sometimes box needs to be placed more caudally on top of the diaphragm, because 1) movement of the box is that small on the rigid area or 2) external diaphragm hides the visibility of the marker. Still, on average only 2% of all patients are such that the chest wall movement is less than 5 mm in vertical direction when they come to linac with AlignRT. On average, chest wall ROI movement in vertical direction from FB to DIBH is 1.1cm. On the first fraction with AlignRT, 2% of the patients are after vertebra match in BH DICOM inside chest wall ROI thresholds, although they are breathing free, mostly indicating that they are completely belly breather without significant movement on the chest wall. Then re-planning is needed and once those patients are instructed to take a breath slightly differently at the CT, BHs and treatment has gone fine.