An automatic alert tool for head and neck patients requiring adaptive radiotherapy

S10 Inside this scenario two important aspects must taking into account: image registration methodology and network facilities. Image registration, e.g. MR-CT, PET-CT, SPECT-CT and serial intramodality registration must be applied in the right way following 3D and 4D approach. Using image registration methods ROI and VOI information planned inside the planning CT must be related with “on board” imaging. Networking facility now are a must inside this new scenario; only through these support images, patient set up, dose planning may be transferred across the different systems (from planning to dose deliveries). Network in Radiotherapy must be sure, fast, integrated with Hospital facilities, standard following the International Standards. The complexity of the technologies used and the more relevant multidisciplinary approach in radiotherapy environments need a new profile of MPE inside the radiation oncology departments. The rule and the profile of MPE for this new scenario will be presented. The MPE is the Qualified Medical Physicist level with accredited clinical training in the area/s of Digital Imaging, Nuclear Medicine or Radiotherapy. His profile needs: Knowledge (facts, principles, theories, practices), Skills, as the ability to use knowledge and know-how to complete tasks and solve problems (both cognitive skills involving the use of logical, intuitive and creative thinking and practical skills involving manual dexterity and the use of methods, materials, tools and instruments) and Interdisciplinary Competence (responsibility and autonomy). Keywords: imaging, radiotherapy, MPE

32 Evaluation of the clinical usefulness for using verification images during frameless radiosurgery T. Gevaert1, M. Boussaer1, B. Engels1, C.F. Litre ´2, A. Prieur3, D. Verellen1, J. D’Haens1, P. Collin3 and M. De Ridder1 1 Department of Radiotherapy, Universitair Ziekenhuis Brussel, Vrije Universiteit Brussel, Brussels, Belgium, 2Department of Neurosurgery, CHU de Reims, Reims, France, 3Department of radiotherapy, CHU de Reims, Reims, France Purpose: This study was designed to evaluate if verification images can correct for mechanical inaccuracy of table rotations during frameless radiosurgery treatment. The added value of these verification images was also analyzed for intrafraction motion reduction. Material and methods: Multiple hidden target tests (HTT) were performed to measure the overall accuracy at different table rotations (0 , 90 and 270 ) with and without stereoscopic x-ray imaging and 6 degree-of-freedom (DOF) positioning and correction for mechanical inaccuracy of the table rotations (previously measured with the Winston-Lutz-test). To analyze the intrafraction motion reduction, sixteen patients with trigeminal neuralgia immobilized with a Brainlab frameless mask system and treated with frameless stereotactic radiosurgery were enrolled. Patient positioning was performed with the Brainlab ExacTrac stereoscopic x-ray system. After every beam delivery and table rotation (10 per patient) verification images and 6DOF registration/positioning were performed. Those data were used to analyze if the intrafraction motion could be reduced. Results: The results of the HTTs showed an overall three-dimensional (3D) accuracy of 0.66mm (SD 0.38mm) for the table at 0 . Rotating the couch to 90 and 270 degrees table position, showed a deviation of the overall 3D accuracy of 0.79mm (SD 0.50mm) and 0.55mm (SD 0.42mm), respectively, comparable to the results of the Winston-Lutz test. Repositioning with the use of stereoscopic xray imaging and 6DOF positioning at 90 and 270 degrees table position, showed a positioning correction up to 0.30mm (SD 0.18mm) and 0.20mm (SD 0.12mm), respectively

SFPM Annual Meeting 2012 During patient treatment, verification images were taken every 2 minutes. The mean 3D intrafraction shift was 0.33mm (SD 0.17 mm). This is a motion reduction of 56% compared to verification imaging only after treatment (0.58mm SD0.43mm). The mean intrafractional rotational errors were comparable for the vertical, longitudinal, and lateral directions: -0.01 (SD 0.54 ), 0.06 (SD 0.41 ), and -0.03 (SD 0.38 ), respectively. Conclusions: With verification images taken during treatment and at different table positions the mechanical inaccuracy of the table rotation can be corrected. This will improve target positioning with respect to treatment isocenter. Furthermore, the 6DOF verification and positioning during treatment will reduce the intrafraction patient motion into the mask immobilization. Keywords: Frameless radiosurgery, thermoplastic mask, verification images

33 Geometric accuracy of real-time tumor tracking with the gimbal Linac system of the novel VERO SBRT system T. Depuydt, T. Gevaert, D. Verellen, K. Poels, M. Duchateau, K. Tournel, T. Reynders and M. De Ridder Department of Radiotherapy, Universitair Ziekenhuis Brussel, Vrije Universiteit Brussel, Brussels, Belgium Purpose: VERO is a novel platform for image guided stereotactic body radiotherapy. Orthogonal gimbals hold the linac-MLC assembly allowing real-time moving tumor tracking. This study determines the geometric accuracy of the tracking. Materials and Methods: To determine the tracking error, a 1D moving phantom produced sinusoidal motion with frequencies up to 30 breaths per minute (bpm). Tumor trajectories of patients were reproduced using a 2D robot and pursued with the gimbals tracking system. Using the moving beam light field and a digital-camerabased detection unit tracking errors, system lag and equivalence of pan/tilt performance were measured and characterized in terms of systematic error, standard deviation and E90% each calculated over a whole number of sinus periods. Results: The systematic errors, calculated as the average of the difference between tracked object and tracking beam position, were below 0.14 mm for all sinus motion sequences. Without forward prediction, the tracking error is strongly dependent on the motion frequency and values of up to 2.16 mm and 2.99 mm for pan and 2.12 mm and 2.98 mm for tilt were seen for standard deviation and E90% respectively. Compensating the system lag with a forward prediction gimbals control of 50 ms reduced both maximum values of standard deviation and E90% respectively to below 0.46 mm and 0.82 mm for pan, and 0.31 mm and 0.59 mm for tilt. For the patient tumor motion tracks, with compensation of the system lag, the tracking error was reduced to a 2D deviation vector modulus E90% of 0.54 mm and standard deviation of 0.22 mm in pan direction and 0.20 mm in tilt direction. For all sequences the systematic components of the tracking error were below 0.10 mm. Conclusions: In terms of dynamic behavior, the gimbaled linac of the VERO system showed to be an excellent approach for providing accurate real-time tumor tracking in radiation therapy Keywords: Real-time tumor tracking, Tracking error, Organ motion

34 An automatic alert tool for head and neck patients requiring adaptive radiotherapy S. Huger1, P. Graff1, V. Marchesi1, J.C. Diaz2, D. Wolf1, D. Peiffert1 and A. Noel1 1 Centre de Recherche en Automatique de Nancy (CRAN - UMR 7039), Universite´ de Lorraine, CNRS, Centre Alexis Vautrin, Nancy, France, 2Dosisoft, Cachan, France

SFPM Annual Meeting 2012

S11

Introduction: Changing anatomy during head and neck radiotherapy can lead to significant dosimetric consequences for organs at risk (OARs) and/or target volumes. Adaptive radiotherapy can compensate for these variations however the decision to modify the treatment is difficult to evaluate. The aim of this work was to develop, in partnership with DosisoftÒ, a simple in-vivo dosimetric alert tool allowing the detection of situations when patients might require an adaptive radiotherapy. Methods and Materials: The dosimetric tool generates a 3D map of two complementary dosimetric information: a dose tolerance variation map and a dose differential map. The tolerance map is calculated from the dose planned on the initial simulation CT. It displays for each voxel of each delineated organ of the CT plan the acceptable dose variations during treatment without exceeding dosimetric constraints: overdose tolerances for the considered volumes, underdose tolerances for the target volumes. Once the true dose was calculated on CBCT image sets, the differential dose map was generated. Dose variations between planned and true dose distribution are mapped for each voxel. Then, using a rigid registration, the delineated volumes on the planned CT are transferred onto the CBCT. Results: The two maps are automatically compared. When a voxel presents a value superior to the corresponding dose variation tolerance, an alert system is activated and the concerned anatomic area is visually indicated to the physician. As a result of a rigid image registration, the uncertainty in structure positioning on CBCT was corrected considering for each voxel of each structure the maximal dose difference in a 5mm diameter sphere. A preliminary study of 10 H&N patients (weekly CBCT) revealed the tool’s aptitude to detect the cases requiring a new treatment plan. One such patient in whom tumour regression resulted in an increase in the dose delivered to the spinal canal beyond 45Gy was detected. Conclusion: This automatic detection tool allows to rapidly decide if a patient needs a process of adaptive radiotherapy, without physician delineation. Keywords: adaptive radiotherapy, head and neck, dosimetric alert tool

study reports the results of the co-operation between in-room imaging and optical tracking for patient positioning in the frame of high-precision particle therapy. Material and Methods: The OTS system is homemade software developed on the Smart 3D Capture software from BTS Italia. The positions of the external markers, assessed by a triangulation of three infrared cameras, were compared with those expected from the simulation tomodensitometry. The accuracy of the patient positioning using the OTS was also compared with the Imaging System (Medcom) of the treatment room for different localizations (Head&Neck and Pelvis). The retrospective data analysis was based on the first patients treated at the CNAO, all immobilized with the help of head or pelvis masks, on which 7 markers (surrounding the isocenter of the beams and visible by the infrared cameras) were fixed for the whole treatment duration. Results: Differences between the OTS and the Imaging System were on average within the expected localization accuracy, as limited by CT image resolution (3mm). On the first treatments, the corrections applied by the Imaging System after the patient positioning using the OTS only were for the most part assessed to be sub-millimetric regarding the translations (0.230.21mm) and sub-gradual regarding the rotations (0.310.33 ). The position of the patient, recorded during every beam, was also estimated by the OTS to verify the patient immobility during the treatment and give an efficiency score of the session. Conclusion: Although the indication of the optical tracking system cannot replace information provided by in-room imaging devices, the reported data show that OTS preliminary correction might greatly support the image-based patient set-up verification and in most cases allows one to reduce the number of radiographs needed for set-up verification. This ultimately leads to the optimization of the time required per session. It also provides a secondary and independent verification system for the patient positioning featuring an intrinsic accuracy of 0.2 millimeter and 0.3 rotation degree. In addition, one should consider that the role of optical tracking can be easily extended as a motion detection device to check for patient immobilization as well as a motion monitoring systems devoted to dynamic tumor tracking techniques. Keywords: tracking, patient setup, particle therapy

35 Reliability of the optical tracking system for patient positioning at the Centro Nazionale di Adroterapia Oncologica based on the first treatments

TOPIC: EFOMP

M. Desplanques, B. Tagaste, A. Pella, M. Riboldi, G. Fontana, R. Orecchia and G. Baroni Unite´ Bio-inge´nieurie - Centro Nazionale di Adroterapia Oncologica de Pavie - Italy, TBMLab e De´partement de Bio-inge´nieurie - Politecnico di Milano - Milan - Italy, Division de Radiation Oncologique - European Institute of Oncology Milan - Italy, De´partement de Science et Technologies Biome´dicales - Universita` di Milano - Milan - Italy Introduction: In the last number of years, the huge increase of IMRT, VMAT and Tomotherapy clinical treatments have been greatly supported by IGRT (Image Guided RadioTherapy) techniques combined with improved accuracy in patient set-up and motion monitoring provided by IR optical tracking devices (Optical Tracking Systems, OTS). The synergy between in-room imaging and optical tracking, in co-operation with highly accurate robotic patient handling represents the concept for patient-set-up which has been implemented at the CNAO (Centro Nazionale di Adroterapia Oncologica). In-room imaging is based on a double oblique X-ray projection system with variable orientation; optical tracking is based on the detection of markers placed directly on patients’ skin or on the immobilization devices. These markers were used as external fiducials during patient positioning and dose delivery. This

INVITED CONFERENCES 36 EFOMP and Medical Physics in Europe P. Sharp EFOMP President, Aberdeen, UK Abstract not available.

37 Education and training requirements within the Medical Physics Expert project C. Caruana Malta Abstract not available.

38 Legal requirements for Medical Physics and for the Medical Physics Expert: a workshop W. Van Der Putten Galway, Ireland Abstract not available.