ESTRO 2024 - Abstract Book

S4170

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

ESTRO 2024

92.58% for LDD. Notably, while most metrics show LDD with a slight edge over E2E, the versatility of LDD is further emphasized by its landmark error of 1.19±1.16mm when executing a reverse trajectory. Regarding computational efficiency, B-spline took an average of 65.34±35.43s, while E2E and LDD processed at faster rates of 18.86±1.81s and 23.07±2.52s, respectively. Comprehensive side-by-side comparisons of different metrics for the three investigated DIRs are shown in Fig2.

Conclusion:

Our novel approach, capitalizing on the power of INR, offers a means to model motion in a continuous manner for intra-fraction motion modelling. By representing the registration process as a continuous flow, we circumvent the limitations inherent to the classic voxel-based representations, particularly the dilemma of discretization. The tangible enhancements are evident: both INR-DIR (E2E and LDD) methods consistently outperformed classic B-splines regarding TRE, MAE, landmark, and Dice coefficients. Remarkably, the registration times were slashed by more than