ESTRO 2024 - Abstract Book

S4774

Physics - Quality assurance and auditing

ESTRO 2024

right ( x ), superior-inferior ( y ), and anterior-posterior ( z ) room coordinate system and UV -panel coordinate system of the MV imager (MVI) applied by the Unity system were used[3].

The phantom was placed with the sphere near the radiation isocentre and irradiated with 12 equidistant fields in a clockwise rotation. The field edges used for radiation field centre (RFC) determination were the MLC leaf side (transverse U -direction) and MLC tip (longitudinal V -direction). A cosine function was fitted to the positions of the sphere and the RFC in the U -direction of the images. The amplitude and the phase angle of the fit together with the pixel size at isocentre distance determine the shift in xz -coordinates to position the sphere at the RFC. The y -shift in the V -direction defined by the MLC tips is sensitive to the MLC calibration. In the y -direction, the linac RFC calibration was used for MR-to-MV calibration.

The phantom was MRI-scanned with T1 weighted scans (1×1×1 mm 3 ) using the body coil. The MR-to-MV isocentre shift was defined as

Δ r = r S,MV - r S,MR ,

where r S,MV is the position of the sphere relative to the linac MV-radiation isocentre and r S,MR is the sphere's position relative to the MRI scanner's isocentre.

Results:

Figure 2 displays the results of the MV radiations on the MVI panel of the two MRI-linacs located at Odense University Hospital (OUH) and Uppsala University Hospital (UUH). In figure 2(a), the sphere is virtually moved to a position without cosine oscillations. Furthermore, the sphere was virtually moved to the RFC for radiation isocentre xz -shift determination. In the V -direction, figure 2(b), the UUH data shows a slight offset, which might be due to offset