radiotherapy departments to detect errors in dose delivery and to provide an overall check of the whole dosimetry procedure and patient setup. Moreover, in several European countries IVD has become mandatory. A variety of detectors, including thermoluminescent dosimeters (TLD), sili- con diodes and metal oxide silicon field-effect transistors (MOSFETs) are currently available for in-vivo dosimetry. These dosimeters are affected by serious drawbacks, such as long readout time and signal fading (TLDs), strong temperature, angular, dose and dose rate dependence, lack of ra- diation hardness, energy dependence (diodes and MOSFETs). Materials and Methods: A novel detector based on a synthetic single crystal diamond was evaluated for in-vivo dosimetry application. The device prototype was designed as a cable-free portable dosimeter to be operated offline. A basic dosimetric characterization was performed under irradiation with a Co-60 reference gamma beam and a 6 MV clinical photon beam from an Elekta Precise linear accelerator. The main features relevant to IVD application were investigated: preirradiation procedure and repeatability, linearity with dose (0.2 to 7 Gy), dose rate dependence (1 to 5 Gy/min), angular (0  to 360  ) and temperature dependence (20 to 40  C). Results: A negligible preirradiation procedure was found, with a maximum deviation form average reading below 1%. Deviations below 0.5% were found for linearity with dose and dose rate dependence analysis. Fading effect was negligible, with a decrease of diamond detector response below 3% for delays separating irradiation and readout up to 30 minutes. A deviation of 4% of diamond response, measured in the Co-60 beam, be- tween 0  and ±180  irradiation direction was found. Finally, a temperature dependence of about 5% was measured in the investigated temperature range. Discussion: The tested diamond prototype showed good dosimetric fea- tures for off-line IVD applications. Such results, together with the high radiation hardness of diamond and its tissue-equivalence, are very promising for the realization of an energy independent high performance diamond based in-vivo dosimeter. A QA PROCEDURE FOR BRACHYTHERAPY TPS EMPLOYING MODEL BASED DOSE CALCULATIONS, BASED ON MONTE CARLO SIMULATION AND END USER ORIENTED TOOLS V. Peppa, E. Pappas, E. Pantelis, P. Papagiannis. Medical Physics Laboratory, Medical School, University of Athens, 75 Mikras Asias, 115 27 Athens, Greece Background: Model-based dose calculation algorithms (MBDCA) have been recently introduced as an option besides TG43 in 192 Ir HDR brachy- therapy treatment planning systems (TPS). The complexity of these algo- rithms challenges quality assurance (QA) programs requiring verification of dose delivery in patient-specific conditions. The aim of this study is to develop a user oriented QA procedure for TPS employing MBDCAs. Materials-Methods: Mathematical phantoms resembling patient geome- tries, as well as a computational model of a water sphere were constructed. These virtual phantoms were converted to DICOM-CT image series to facili- tate their import to two commercially available systems(BrachyVision v.10.0.33 and OncentraBrachy v.4.4). HDR 192 Ir sources were selected from the TPS source libraries to perform treatment plans. A catheter with a single source dwell position was located at the center of the water sphere, as well as 5 and 10 cm eccentrically. Two plans were created for patient phantoms, one with a multiple source dwell configuration and one where this was replaced by one dwell position at the centroid of the source distribution. In all cases, dosimetric calculations were performed using the model based dose calcu- lation algorithm of the TPS. Monte Carlo (MC) simulation was performed to obtain reference dose distributions using BrachyGuide, a brachytherapy dedicated software tool prepared in-house for the automatic preparation of MCNP input files from treatment plans exported in DICOM-RT format. A dedicated custom soft- ware suite was also used for the comparison of reference MC data and corresponding TPS results in terms of colormap representation of per- centage differences, Dose Volume Histograms and plan quality indices. Results: Comparisons of MC and TPS results showed a good agreement in the majority of points. Considerable differences were also observed and explained by MBDCA assumptions and optimization settings. These results have to be taken into account in MBDCA commissioning procedures. Discussion: The QA procedure and software tools presented herein may serve as a basis for developing future standards toward providing a quality management of MBDCA-based treatment planning systems. Acknowledgement: Research co-financed by the EU ESF and Greek na- tional funds.NSRF operational Program “Education and Lifelong Learning Investing in knowledge society”, Aristeia I STEREOTACTIC FRAME INDUCED GEOMETRIC DISTORTIONS IN MR IMAGES EMPLOYED IN GAMMA KNIFE RADIOSURGERY APPLICATIONS E. Pappas a , A. Moutsatsos a , P. Karaiskos a, b , E. Pantelis a , E. Georgiou a , M. Torrens b , I. Seimenis c . a Medical Physics Laboratory, Medical School, University of Athens, 75 Mikras Asias, 11527 Athens, Greece; b Gamma Knife Department, Hygeia Hospital, Kifisias Avenue and 4 Erythrou Stavrou, Marousi, 15123 Athens, Greece; c Medical Physics Laboratory, Medical School, Democritus University of Thrace, 2nd building of Preclinical Section, University Campus, 68100 Alexandroupolis, Greece Purpose: To asses geometric uncertainties present in MR images neces- sitated for target localization and delineation in Gamma Knife radio- surgery. Specifically, this work focuses on the geometric distortions induced by eddy currents associated with the Ti-based Leksell stereotactic frame used for patient immobilization. Materials and Methods: A plexiglass phantom -recently developed by our group- was utilized. It encompasses 947 3mm-diameter holes interspersed in the 3D space which serve as control points for distortion detection and eval- uation. Control point distribution pattern ensures that the entire stereotactic space is monitored. The phantom accurately fits the Leksell localization box and stereotactic frame, but can also be fixed without the latter. The reversed read gradient technique, along with a reference CT scan, were employed to evaluate geometric distortions in the phantom imaged at 1.5T with clinically- used gradient echo pulse sequences (axial 3D-T1w and coronal balanced- T2w). An in-house, semi-automatic MATLAB software was developed to assist control point localization and distortion evaluation. Derived 2D distortion maps were compared with corresponding ones deduced without the frame. Results: MRI-related distortions without the presence of the frame were limited to 1mm. Contrarily, severe frame-induced distortion was detected in regions neighboring the frame base. In T1w images, this distortion exceeded 5mm in a 20mm radial distance from the frame. Distortion decreased rapidly with the distance from the frame and obliterated at about 70mm. The control point disposition was always directed towards the frame center and did not change sign relatively to the read gradient polarity. Slightly larger distortion was detected with the frequency encoding direction being perpendicular to the most proximal side of the frame base. In coronal balanced-T2w images, the frame evidently distorted the nearside fiducials. Conclusion: Geometric inaccuracies induced by the Leksell stereotactic frame should be considered in GK radiosurgery applications with targets lying in the vicinity of the frame base. Total distortion inherent in MR images might reach inacceptable levels in extracranial cases, such as cer- vical spine, head and neck tumours (applicable in Perfexion model), where the target is close to the frame base. SYNTHETIC SINGLE CRYSTAL DIAMOND DIODE INCLINICAL DOSIMETRY OF HIGH DOSE PER PULSE ELECTRON BEAMS FOR INTRAOPERATIVE RADIATION THERAPY (IORT) Marco Marinelli a , G. Verona-Rinati a , M.D. Falco b , M. Pimpinella c , S. De Stefano d , A. Ciccotelli d , G. Felici d , F. Marangoni d . a INFNeDepartment of Industrial Engineering, University of Rome “Tor Vergata”, Italy; b Tor Vergata University General Hospital, Roma, Italy; c ENEA-INMRI, Istituto Nazionale di Metrologia delle Radiazioni Ionizzanti, Roma, Italy; d S.I.T. - Sordina IORT Technologies spa, Italy Background: Accurate clinical dosimetry of electron beams produced by special linear accelerators dedicated to intraoperative radiation therapy (IORT) is challenging due to the presence of very high dose per pulse (1-10 cGy per pulse) with respect to conventional accelerators (<0.1 cGy per pulse). This high dose rate can affect the dose determination by means of dosimeters currently used for absolute and relative dosimetry (e.g. ionization chambers). In addi- tion, due to the high dose gradients of electron beams, small volume dosim- eters are needed. Recently, the PTW microDiamond type 60019 has been Abstracts / Physica Medica 30 (2014) e45ee74 e68