ESTRO 2021 Abstract Book

S1485

ESTRO 2021

without the ribcage in the phantom.

Results The intensity line profiles across the tumour (Fig. 1a) of the original and bone suppressed X-ray image regions (Fig. 1b+c) show a strong reduction of the bone intensity (represented by the almost flat red line outside of the tumour section) compared to the original X-ray image (represented by blue line with high peak plateaus caused by ribs). A reduction of bone/lung tissue contrast median values between original and bone suppressed images was from 0.0945 to 0.0023. (Fig. 2b). The tumour/lung contrast remained constant (Fig. 2a). TMM succeeded in all cases when ribs were suppressed, but failed when original X-ray images were used. The registration error of the tumour trajectory between ground truth and bone suppressed images was found to be 1.52 mm for RMSE and 1.27 mm for MAE.

Conclusion The proposed bone suppression method allowed accurate TMM when the tumour was eclipsed by the ribs in an anthropomorphic lung phantom. These promising results need to be validated with a patient study to evaluate benefits in clinical practice. PO-1759 Pre-treatment CBCT rotations as a predictor of intra-fraction motion in cranial SRS patients. R. Dawson 1 , R. Brass 1 1 The Clatterbridge Cancer Centre NHS Foundation Trust, Physics Department, Liverpool, United Kingdom Purpose or Objective Intra-fraction motion is a concern for all radiotherapy patients, but especially radiosurgery patients due to the significantly reduced PTV margin and delivery in a single fraction increasing the impact of random errors. The