ESTRO 2024 - Abstract Book

S3349

Physics - Detectors, dose measurement and phantoms

ESTRO 2024

Results:

All machine-related errors that could be detected by the pre-treatment verification were also detected by the 2D IVD system. Neither pre-treatment verification nor in vivo dosimetry could detect: MLC banks retracted by 0.5mm (first arc); both MLC banks moved in by 0.5mm (both arcs) and collimator ±1º rotations (both arcs).

In relation to patient-errors, the 2D IVD system only detected the 7 mm longitudinal shift (both arcs) and the change in phantom thickness of 2 cm (second arc).

Image 2. Local Gamma Evaluation of pre-treatment verification and IVD of the errors introduced. In red, the cases in which the error has been successfully detected.

Conclusion:

EPID-based 2D IVD software detects the same machine related errors as pre-treatment verifications.

However, transit IVD is not sensitive enough to detect errors up to 7 mm in patient’s setup and thickness changes smaller than 2 cm for the studied VMAT spine SBRT treatment. If we increase the acceptance criteria from 95% to 98%, we would detect a change of diameter of 1cm and all the longitudinal shifts; lateral and vertical shifts would remain undetected.

The sensitivity of transit in vivo dosimetry to detect positioning errors depends on the treatment technique and on the treatment site, being less sensitive for complete arcs and when the treated region is more homogeneous.

Keywords: In-vivo dosimetry (IVD), EPID, error detection