84 A NEW VERIFICATION METHOD OF DOSE DELIVERY FOR PROTON STEREOTACTIC RADIOSURGERY. SM. Vatnitsky, R.W.M. Schulte, R. Galindo, H.J. Meinass, and D.W. Miller Loma Linda University Medical Center, Loma Linda, USA The verification of the dose delivery and the study of spatial and dosimetrical accuracy ofproton radiosurgery (PRS) with multiple beams is limited due to the lack of sufficient detector systems. In the present paper we report the technique and results of a new verification method for PRS ofintracranial targets using a nigh sensitivity radiochromic film. Calibration of the film and single proton beam characterization were performed with unmodulated and modulated beams. Radiochromic film was found to have lower sensitivity at the distal edge compared with an ionization chamber. However, the penetration of the beam at the level of 80 - 50% range and lateral profiles measured with both film and ionization chamber were in good agreement. The planning steps of PRS, i.e.: CT scan, target delineation, design of a treatment beams, aperture shapes, and tissue compensators were simulated using a water phantom with embedded objects to represent a target, alignment fiducial markers, and tissue inhomogeinity. During the planning CT scan and treatment the phantom was attached to a stereotactic flame. X-ray films and digitally reconstructed radiographs were used for phantom alignment. Dose distributions delivered to the target during PRS with five non-coplanar proton beams were measured using a radiochromic film. A comparison of measured and calculated dose distributions demonstrated agreement within margins and uncertainties usually employed for treatment planning of patients. As the simulated procedure was the same as during a patient treatment, the combined uncertainty of the dose delivery for the reported phantom simulation should be close to that wnich can be expected for a patient treatment. 86 3-D PLANNING IN INTERSTITIAL BRACHYTHERAPY IN THE HEAD AND NECK: ACCURACY OF MATCHING IMPLANT GEOMETRY BASED ON LOCALIZATION RADIOGRAPHS WITH CT A.G. Visser, I.K.K. Kolkman-Deurloo, P.P. Jansen, P.C. Levendag Dr. Daniel Den Hoed Cancer Center/Academic Hospital Rotterdam, The Netherlands Background and purpose. In clinical applications of brachytherapy, the use of 3D information from CT/MRI imaging is applied much less than in external beam RT, possibly because of the custom to position applicators/sources "a rue" end to localize the implant geometry afterwards with a pair of radiographs. There is, however, a need for 3D soft tissue information in brachytherapy, for target volume definition and dose-volume analysis both for tumor and critical normal tissues. Stereotactic frames are applied widely in brain implants, but are considered impractical for other body sites, Material and methods. If implant reconstruction for planning for whatever reason cannot be done from CT images with the catheters in place, there is a need for matching the coordinate systems of CT and the localization radiographs. Different matching methods have been tested for flexible catheter implants of the tonsillar region and the soft palate. In order to evaluate the accuracy, CT-images with the catheters implanted have been used as benchmark for comparison with reconstruction from a pair of radiographs, followed by image registration. Preliminary results. Two methods have been developed and tested: i) the indication of internal markers (bony landmarks) and/or external markers in both the CT and the localization images followed by a (Simplex) optimization, and ii) matching of CT-derived DRR's (digitally reconstructed radiographs) with the localization radiographs. Displacement of implanted catheters, due m anatomical changes (oedama and/or patient position) during the interval between radiographs and CT, appears to have a noticeable effect, limiting the accuracy in some cases to about 2 ram. The matching procedures can have a similar or bsttter accuracy. Results of matching by means of markers depend on the number of markers used and the CT slice thickness; if sufficient markers are available, the accuracy can be derived from the remaining markers not used for matching. The different image registration methods and their accuracy limitations will be discussed, illustrated with clinical examples. 85 INTERSTITAL BRACHYTHERAPY USING COMPUTER ASSISTED NAVIGATION E. Hensler', Th. Auer', O. Gaberb, A. Gunkel¢, W. Freysinger c, R. Bale=, W. Thumfart c and P. Lukas' aDepartment of Radiotherapy and Radiooncology, blnstitute of Anatomy, CDepartment of ENT, Irmsbruck Austria 1. PURPOSE This paper describes the adaptation of computer assisted navigation to interstitial brachytherapy. The use of computer assisted guidance supports planning the entrance path of the needles to avoid injury of vessels and to achieve an optimal localization in relation to the tumor. The improved positioning accuracy allows treatment with smaller margins around the tumor. 2. MATERIALS AND METHODS A LED based system (EasyGuide, Philips) and a mechanical system (ESG Viewing Wand) were in use. For checking the accuracy two phantoms were designed to simulate human tissue. To find the best setup investigations with an anatomical specimen and the Alderson phantom were performed. Up to now two patients got a fractionated therapy (5x2 Gy) with the aid of this technique. 3. RESULTS The standard deviation was 2.9 mm (maximum 11.9 mm) at 115 attempts. With an more rigid table a standard deviation of 1.1 ram (maximum 5.2 ram) at 51 attempts has been obtained. This accuracy is sufficient for the needs of interstitial brachytherapy. 4. CONCLUSIONS The method presented in this paper may lead to substantial improvements in interstitial brachytherapy, especially reduction of side effects. The next steps will be to gather data in the current clinical trial and to improve the technique. 87 IORT-NOVAC 7 A NEW LINEAR ACCELERATOR FOR ELECTRON BEAM THERAPY M. Fantini, F. Santoni, A. Soriani, 0. Creton Hitesys SpA. Apt-ilia, Italy ; M Benassi, L. Begnozzi, Laboratorio di Fisica Medica C.R.S. Istituto Regina Elena, Rome, Italy. The new linear accelerator for beam therapy, named NOVAC7, is described. This a.pparatus was specifically designed and realized by ENEA (Italian National Agency for New Technolo- gies, Energy and the Envh'onment) and HITESYS S.p.A. for the lntra Operative Radio Therapy (IORT) treatments. Up to now, radioprotection requirements have confined the use of linear accelerators inside the bunker of the radiotherapy departments. Wider use of intraoperative electron beam therapy has been limited by the logistics, costs and hazards of moving the anesthetized patient from the Surgical Suite to the Radiation Oncology Department. The NOVAC7, a new and more advanced equipment handled by a robotic manipulator, greatly improves this methodology. In fact it allows carrying out the radiation treatment inside the operating room and two ways of irradiation: the traditional one and by a mechanical scanning of a small beam. The nominal beam energy of NOVAC7 is variable to the maximum of 9 MeV and the scanned field size can be at maximum 12X12 cm~. Measuremeants to characterize the small beams and their composition have been performed with Gafchromic MD-55 f~ns. Experimental results about stray- radiation and dosimetric characterization are presented. Radiotherapy and Oncology, Volume 40, Supplement 1, 1996, Page S24