ESTRO 38 Abstract book

S245 ESTRO 38

Purpose or Objective To report on an upgrade of the Anthropomorphic Dynamic breAthing Model (ADAM) phantom for QA of real time respiratory tracking systems. Material and Methods ADAM is a 3D printed human torso with realistic embedded ribs and spinal cord, a moving upper chest wall and simulated lungs. An Arduino programmable board, integrated in the phantom body, drives movements of lung along linear or elliptical paths and of upper chest wall up and down. Linear paths dephasing thorax motion and signals based on real patient breathing traces are implemented. In the current version a new Tracking Tool (TT) (fig 1A) has been realized to enable accuracy tests of tracking systems based on fiducial-less tumor detection. The TT consists in four blocks, each hosting a quarter of sphere, simulating a lung-tumor structure, built using a 3D printer and different filling percentage This results in X rays images with realistic contrast between the target and the surrounding material. The TT can host two small films and can be inserted in any position inside the phantom thorax enabling to simulate different tumor detection uncertainties, as observed in real patient cases (fig 1 B). The TT is moved by the same system used to move lungs.

Conclusion From the above results, one notices that highly inhomogeneous out-of-field distributions in the patient can easily be achieved for different collimator rotations during VMAT irradiations. Extra precautions can be taken into consideration when treating patients with HD120MLC or other devices presenting the same behavior. To avoid directing the extra-focal radiation into a radiosensitive organ during VMAT treatment, the collimator angle can be optimized as described to avoid sensitive organs at risk. This issue is of major importance in pediatric patients. [1] Ghareeb, F., et al., Physica Medica, 2018. 52: p. 57-58 PV-0478 A new tool to test tracking systems based on tumour detection S. Pallotta 1 , S. Calusi 2 , L. Masi 2 , L. Marrazzo 3 , C. Talamonti 1 , L. Livi 4 , G. Simontacchi 4 , L. Foggi 5 , R. Lisci 6 1 University of Florence- Azienda Ospedaliero Universitaria Careggi, Department of Biomedical Experimental and Clinical Sciences "Mario Serio"- Medical Physics Unit, Florence, Italy ; 2 IFCA, Medical Physics and Radiation Oncology, Florence, Italy ; 3 Azienda Ospedaliero Universitaria Careggi, Medical Physics Unit, Florence, Italy ; 4 University of Florence- Azienda Ospedaliero Universitaria Careggi, Department of Biomedical Experimental and Clinical Sciences "Mario Serio"- Radiotherapy Unit, Florence, Italy ; 5 University of Florence, Department of Biomedical Experimental and Clinical Sciences "Mario Serio", Florence, Italy ; 6 University of Florence, of Agricultural- Food and Forestry System, Florence, Italy

ADAM and the new TT were used to perform end-to-end tests of a CyberKnife® System (Accuray,USA). An isocentric plan was created targeting the beams on the TT. Two orthogonal EBT3 films were inserted in the sphere and the plan was delivered while moving the sphere and the thorax according to: Linear, Elliptical, Linear no sync (thorax and TT motion not always synchronized), Linear patient (based on a real patient breathing) and Linear patient no tracking (with the tracking system deactivated) respiratory traces. The same plan was delivered also in static condition. Films were analyzed using an E2E dedicated software (Accuray, USA), which provides the distance between the centroid of planned and delivered dose distributions as global tracking error. Films irradiated in static and moving conditions were compared with FilmQA™ Pro (Ashland Inc.,USA). Results The TT was detected by the tracking algorithm with an uncertainty (range 10%-20%) close to what observed in real patient treatments (mean 15.5%; range 5%-32). In one of the two projections, uncertainty was higher (15%-20%) due to the superimposition of a radiopaque heart-like structure (fig 1C). Tracking errors derived by the E2E software were all below 0.95 mm (range 0.6-0.9 mm). The maximum difference in tracking error between two repeated tests was 0.4 mm. In table 1 gamma passing rates obtained by comparing films acquired in dynamic and static conditions are reported. In particular GI 3%/1mm (local criteria) pass-rates scored above 90% for all respiratory