ESTRO 2021 Abstract Book

S1492

ESTRO 2021

Figure 2. Rotational corrections in the three rotational axes obtained with the Catalyst and XVI systems.

Conclusion There was a good agreement between the rotational setup corrections determined with the Catalyst system and the corrections determined in the CBCT-CT image registration. The good agreement demonstrates the high accuracy of the surface scanning system in the positioning of patients prior to RT of breast cancer. PO-1765 Markerless 3D tumor tracking during single-fraction free-breathing flattening-filter-free lung SBRT I. Remmerts de Vries 1 , M. Dahele 1 , H. Mostafavi 2 , B. Slotman 1 , W. Verbakel 1 1 AmsterdamUMC, Radiation oncology, Amsterdam, The Netherlands; 2 Varian, Medical Systems, Palo Alto, California, USA Purpose or Objective Single fraction lung SBRT is gaining interest. Tumors may move differently during treatment than they do on the planning 4DCT scan and positional verification during irradiation would increase confidence and reduce the chance of geographic miss. The ability to acquire planar 2D kV images during VMAT is already clinically available on current commercial linear accelerators, and so markerless tracking based on these could offer a widely available and low-cost verification solution. We evaluated template matching and triangulation for 3D positional verification on fluoroscopic kV images acquired at 7 frames/second during single fraction, 34Gy, flattening filter free (FFF) VMAT lung SBRT of small lung lesions. Materials and Methods Tumor tracking was performed during irradiation in 6 consecutive patients (7 lesions). treated on a TrueBeam® system (version 2.7; Varian Medical Systems, Palo Alto, CA). VMAT was delivered using 3-6 arcs. Positional verification was performed using CBCT prior to the 1 st arc and after every 2 arcs. Online tracking, using non-clinical software (not able to control the treatment unit), consisted of template matching of each kV image followed by triangulation with previous (>20 degrees) matches. It was retrospectively analyzed how often the tumor was within the 5mm clinical PTV margin and how often it would have been within a 3mm margin. The times for imaging and delivery sub-steps, from the start of 2D kV images (performed after laser- based patient positioning) until the end of the last arc, are reported. Results Tumor diameter was 8-14mm. Longitudinal motion on 4D planning CT was 1-8mm. For all 7 lesions combined, 3D tumor position could be determined for, on average, 71% of the total irradiation time (range 51%-84%). During the tracked period, the longitudinal, lateral and vertical position of the tumor was within the clinical PTV (ITV + isotropic 5mm margin) 98.8% of the time. If the margin had been 3mm, it would have been within this 93.4% of the time (see figure below). The total time from first set-up imaging to the end of the last arc ranged from 16.0 minutes to 21.4 minutes (mean 18.0 minutes, SD 2.0). The average irradiation (beam-on) time needed to deliver all arcs was 5.4 minutes (SD 1.0). Mean arc length was 178 degrees (SD 45.0).