ESTRO 2021 Abstract Book

S1470

ESTRO 2021

Fig 1 Beam overrides in February 2021 71 fields were irradiated in QA mode by non-physics staff. (Technologists using therapist's credentials) Incomplete site setups and missing documents were found. These errors were corrected but not quantified. Neither changes between non-twin machines nor privilege escalation were detected during the evaluated period. Conclusion We found that in our clinic table overrides are a common practice, occurring daily with high frequency. This has prompt an analysis on our couch tolerances and a survey of actions in order to reduce the number of overrides. Besides, a reassignment of Mosaiq permissions among RT workers is currently under evaluation. PO-1747 Tracking accuracy and margins for synchronized deliveries on Radixact using patient-specific motions W. Ferris 1 , W. Culberson 1 , J. Bayouth 2 1 University of Wisconsin Madison, Medical Physics, Madison, Wisconsin, USA; 2 University of Wisconsin Madison, Human Oncology, Madison, Wisconsin, USA Purpose or Objective The Radixact is capable of delivering tomotherapy treatments with target tracking, called Synchrony. The target location is modeled during treatment using kV radiographs and LEDs on the patient’s chest. The collimation is steered in real time using jaw sway and MLC shifts. Analyses of Synchrony tracking have been restricted to measurements with mathematically-derived phantom motion. The purpose of this work was to quantify tracking accuracy of Synchrony for realistic patient motion and to investigate target margins required to account for tracking uncertainties. Materials and Methods Motion traces were derived using 4DCT datasets. The location of the target and the chest LED were observed on the extreme phases of the 4DCT, to be used to scale phantom motion in each direction. The 4DCT continuous respiratory amplitude pattern was used as the motion pattern for the target (X/Y/Z) and the patient’s chest (Z), only scaled by different magnitudes. In this manner perfect correlation between internal and external motion was assumed for the initial investigations. A 3-D motion control stage was used to drive the Delta4 to mimic 3D tumor motion and a separate chest motion stage was used to mimic chest and LED motion. Treatments were delivered with Synchrony tracking the motion of four gold fiducials in the Delta4. Log files of the modeled tumor position throughout treatment were compared to the known phantom motion. The RMS error between the two throughout treatment is denoted δ rms and that with setup/registration errors removed is denoted δ rms_adj . Digital Poster: Intra-fraction motion management