ESTRO 2024 - Abstract Book

S4238

Physics - Intra-fraction motion management and real-time adaptive radiotherapy

ESTRO 2024

exhale phase template cine images, the respiratory mode using mid-position phase template cine images, and the breath-hold mode using end-inhale phase template cine images.

The MM performance was quantified using a through-plane motion quality factor (QF) and jitter QF. The former represents the ratio between the target area in the template and 2D cine MRI, while the latter indicates frequent changes in estimated cranial-caudal motion. Furthermore, the motion amplitude derived from the real-time motion signals were compared for the different 3D reference volumes. Individual breathing cycles were extracted by identifying the end-inhale locations, and the average motion amplitude of all breathing cycles was determined. For the SMS-4D-MRI data, the translation between the end inhale and end-exhale phases was manually determined and compared to the 1D-RNAV motion amplitude.

Results:

Figure 1 shows anatomical examples of vendor-provided 3D-MRI scans and high-definition motion-compensated 4D MRI derived images in the end-exhale, mid-position, and end-inhale phase. Similar motion signals were obtained with the 4D-MRI derived 3D volumes as reference for real-time MM as compared to the vendor-provided 3D scans (Figure 2A). The average (min:max) real-time motion amplitude differed 0.5 (0.2:1.1) mm, 0.4 (0.1:0.8) mm, and 0.5 (0.1:1.5) mm in respectively left-right, anterior-posterior, and cranial-caudal directions for the different 3D reference volumes. The through-plane motion QF was comparable between the different simulations (Figure 2B). The differences of the median amount of jitter were 0.0—0.4 mm between the 3D references volumes, and the median values were 0.84, 0.86, and 1.18 mm for the exhale, respiratory, and breath-hold modes, respectively (Figure 2C). A jitter value larger than 5 mm (indicating MM failure) was observed in 0.04% of the registrations. The motion amplitudes derived from SMS-4D-MRIs and their 1D-RNAV signals agreed within 1.1—3.0 mm.

Conclusion:

Utilizing 4D-MRI data acquired during free-breathing enables the reconstruction of improved motion-compensated daily MRIs in a user-selected respiratory phase for plan adaptation on the MR-linac without compromising the quality of MM used for gated deliveries.