Abstract Book

ESTRO 37

S560

We have chosen to analyze cathode pulse voltage (VKAT), magnetron current , magnetron electromagnet current, thyratron reverse current (IINV) and temperature. Four different models have been fitted, one per energy. Results The SVM has a binary output which corresponds to whether the input data is contained in the learned frontier or not. Several parameters have been used which affect directly the effectiveness of the system. Overfitting the model will lead to false error detections, while underfitting renders the model insensitνive. Parameter ν in scikit-learn SVM implementation takes into account the fraction of training data that will be considered out of the boundaries and γ will transform the frontier to better fit the data. Some examples in a simplified 2D model can be shown in figure 2. Conclusion

contoured using automatic tools to assure consistency among the different sets of contours. We then performed DIR among the nine CT series using the version of ANACONDA implemented in RaySt ation v5. This is a hybrid algorithm in which information from both the voxel values and the distance between homologous designated landmarks (ROI/POI) may be used. As a measure of image similarity the correlation coefficient is employed. We have analyzed the three different methods of DIR provided by the algorithm: first using only image information, second using only ROI, and third the full method using both types of information. To assess the accuracy of each DIR, we used the Dice similarity coefficient (DSC) of the rectal ballo on, as well as the target registration error (TRE) of the markers. To evaluate the invertibility, we doubly deformed the rectal balloon and markers, from one imageset to the other and then back to the original imageset, and measured again the DSC and TRE, and the change of volume of the balloon.

Results

It is indeed possible to build a QA system for a mobile IORT accelerator which is able to detect and/or anticipate linac failures. Care has to be taken selecting the parameters ν and γ of the model to avoid over or underfitting. It is important to train the model with a reasonable amount of data to properly calculate the anomaly frontier. PO-1004 Assessment Of A Commercially Available Algorithm For Deformable Image Registration M.A. De La Casa 1 , D. Zucca 1 , J. García 1 , J. Martí 1 , P. Fernández-Letón 1 1 Hospital Universitario Madrid Sanchinarro - Grupo Hospital de Madrid, Radiofísica y Protección Radiológica, Madrid, Spain Purpose or Objective The aim of this work is to evaluate the accuracy and invertibility of ANACONDA, one of the algorithms for deformable image registration (DIR) bundled in the commercial treatment planning system RayStation. Material and Methods We designed a home-made phantom using a rectal balloon and the compact bone and trabecular bone inserts from the CIRS phantom. We attached radiopaque markers to the surface of the balloon. These elements were submerged in a water tank in a position similar to that of the organs in the pelvic region. We acquired nine CT image series of the phantom varying the filling of the rectal balloon. The filling varied in volume, with 50, 70 and 90 cm 3 , as well as in composition, with three different mixtures of water and iodinated contrast: 20% of contrast, 5% and 1%. The image series were acquired with slice thickness 1.5 mm. In each imageset the balloon and the bone inserts were

The most important results of our study are summarized in the attached table: a) the reduction in the distance between homologous markers, with a TRE of 2-3 mm if ROI information is used; b) a DSC of at least 0.98 is obtained when ROI information is used, obtaining bad results only when no ROI are used and the contrast between the balloon and the medium is low. The displacements applied on the markers by the algorithm are invertible, within an uncertainty of 2 mm, if ROI information is used as input for the algorithm. Conclusion ANACONDA achieves good accuracy and invertibility for the kind and amplitude of the deformations present in our phantom, especially when good and consistent ROI information is used. In this case, the accuracy and invertibility are within 2-3 mm. It is inadequate to use image information only, but in the highest contrast cases. Moreover, the deformation field around the balloon does not affect the shape of nearby structures, such as the bone inserts, which preserved their shape and volume after the registration. PO-1005 Towards Dutch consensus on treatment planning for breast cancer, current variation across institutes C. Van Beveren 1 , N. Bijker 2 , L.J. Boersma 1 , K.F. Crama 2 , C.W. Hurkmans 3 , T.T. Nuver 4 , M.J.C. Van der Sangen 3 1 MAASTRO Clinic, Radiation Oncology, Maastricht, The