Abstract Book

S1229

ESTRO 37

control probability (TCP) and also makes possible to switch to the hypofractionated treatment, which leads to multiple benefits, i.e. the optimization of resources, lowered costs, and the patients' improved quality of life.

Electronic Poster: RTT track: Imaging acquisition and registration, OAR and target definition

EP-2348 MRI acquisition in radiotherapy treatment position: sequence optimization for head and neck cancer A. Largent 1 , L. Marage 1 , I. Gicquiau 1 , G. Gambarota 1 , H. Chajon 2 , O. Acosta 1 , J.C. Nunes 1 , R. De Crevoisier 2 , H. Saint-Jalmes 1 1 University of Rennes 1- INSERM- UMR 1099, Laboratoire du traitement du signal et de l'image LTSI, Rennes, France 2 Centre Eugène Marquis, Radiotherapy department, Rennes, France Purpose or Objective The use of MRI in radiotherapy treatment planning (RTP) for head and neck cancers requires acquisition in RT position with a RTP system (flat table, thermoplastic mask, and head board). Nowadays, this system does not fit with the standard MRI receiving coils. Specific receiving coils for RTP were designed by manufacturers, but at expense of image quality. In this context, this study aims to describe our sequence optimization method designed to obtain fast and reliable MR images for RTP. Material and Methods A water test-object and three volunteers, in treatment position with thermoplastic masks, were scanned. MR imaging was performed at 1.5 T (Optima 450W, GEM, Milwaukee, USA) using the GEM RT open head & neck Suite coil system. Three sets of parameters for isotropic 3D T1 gradient echo sequence and four sets of parameters for isotropic 3D T2 fast spin echo sequence were tested (Table 1). Each image was acquired twice with five repetitions (Fig 1). This double acquisition allows for reliable SNR computation even with parallel imaging acceleration. Except for the v4 T2, all the acquisitions were performed in sagittal direction and then reconstructed in axial to minimize the acquisition time and avoid aliasing. The v4 T2 MRI were acquired directly in axial. To quantify the image quality and find the best parameters quantitative metrics were used. Shannon entropy was firstly computed on the phantom images. Then, SNR and CNR were extracted from the differences of successive volunteer images. In addition, a subjective metric called the Mean Opinion Score, was proposed and obtained from 8 radiotherapists and 3 physicists for volunteer T1 images. Results For the phantom, the minimum entropy value found is 2.05 bits in v3 T1. For all volunteers, SNR and CNR give better results for v3 T1 (Table 1). The native v4 T2 axial images have better resolution and contrast compared to the T2 sagittal acquisitions (Fig 1). The maximum Mean Opinion Scores obtained is for v3 T1 (label score: good). In total, v3 T1 and v4 T2 sets of parameters provide the best image quality.

Conclusion We obtained suitable 3D T1 and T2 images for RTP within a protocol time of 15 minutes at 1.5 T. Our institution will propose this imaging protocol for dose calculation for patients suffering from cancers of the head or neck. EP-2349 Quantification and analysis of coronary artery major ostia motion based on ECG-gated 4D-CT Q. Li 1,2 , Y. Tong 1,2 , G. Gong 1 , Y. Yin 1 1 Shandong Cancer Hospital Affiliated to Shandong University, Department of Radiation Oncology, Ji'nan, China 2 University of South China, School of Nuclear Science and Technology, Hengyang, China Purpose or Objective To quantify and analyze the motion of coronary artery (CA) major ostia which held high risk of radiation-induced injury. Material and Methods 37 female volunteers were selected into our trials, all volunteers underwent intravenous contrast electrocardiography gated four-dimensional computed tomography scanning with thinkness 0.5mm in inspiration breath hold state, which were sorted into 20 phases. The left main coronary artery (LM), the obtuse marginal (OM), the first diagonal branch (D 1 ), the second diagonal branch (D 2 ) ostium and the left anterior descending branch which close to the apex (APX) in the left CA, the first right ventricular artery (V) and the acute marginal (AM) ostium in the right CA according to CA segmentation criteria of American College of Cardiology. Four adjacent slices of all ostia from 0% to 95% phase in cardiac cycle were contoured. The CA major ostia motion was quantified by calculating mean displacement in three co-ordinates: Left-Right (X), Anterio-Posterior (Y) and Cranio-caudal (Z). Using the centroid coordinates of cardiac systolic phase as standard phase, we calculated the displacement of all structures refer to the standard phase. The extensional margin was calculated by the formula: Results (1) The centroid coordinates displacement of left CA ostia for X, Y, Z three coordinates were: LM 7.5±2.6mm, 6.2±2.3 mm, 6.2±2.0mm, OM 11.8±4.0mm, 10.8±4.1mm, 11.8±5.5mm, D 1 5.7±1.9mm, 6.5±4.4mm, 6.7±2.1mm, D 2 5.3±1.9mm, 6.3±3.6mm, 7.1±2.5mm, APX 6.4±3.4mm, 4.6±2.4mm, 7.7±3.3mm. The centroid coordinates displacement of right CA ostia for X, Y, Z three coordinates were: V 15.4±4.0mm, 18.4±4.9mm, 17.4±6.1mm, AM 15.4±3.9mm, 21.8±4.8mm, 18.7±6.2mm. (2) Left CA ostia extensional margin for X, Y, Z three coordinates caused by heart beat were: LM 4mm,3mm,4mm, OM 7mm, 7mm, 7mm, D 1 and D 2 4mm,