Abstract Book

S274

ESTRO 37

Proffered Papers: RTT 5: Motion management and adative stratgies

OC-0520 Positional errors between actual and predicted pancreas based on the surrogate signal using MRI K. Fujimoto 1,2 , T. Shiinoki 3 , Y. Yuasa 1,2 , M. Yamane 2 , H. Hanazawa 3 , K. Shibuya 3 1 YAMAGUCHI UNIVERSITY, Radiation Oncology- Graduate School of Medicine, UBE, Japan 2 YAMAGUCHI UNIVERSITY, Radiological Technology, UBE, Japan 3 YAMAGUCHI UNIVERSITY, Radiation Oncology, UBE, Japan Purpose or Objective To manage the respiratory motion for radiotherapy, tumor position prediction based on surrogate signals is commonly used because it is assumed that there is a strong correlation between internal tumor motion and external surrogate signal. However, some studies reported that the tumor border position does not correlate well with abdominal wall or diaphragmatic position. Few study reported that the relationship between internal motion of each pancreatic region and the external surrogate signal. We aimed to quantify respiratory motion of each pancreatic region (head, body, and tail) and to evaluate the predicted positional error of pancreas when the abdominal wall motion or diaphragmatic motion were used as the surrogate. Material and Methods The evaluation method is shown in Fig.1. 2D cine-MRI images of the pancreas were obtained in 10 heal thy volunteers during free-breathing with audio couching. MRI images were obtained for 20 seconds at 0.33 second intervals in sagittal plane using a balanced steady-state free precession sequence (TrueFISP) on a 3T scanner (MAGNETOM Prisma; SIEMENS). To characterize respiratory motion of each pancreatic region and its correlation with the surrogate signals, 2D template images of the pancreas, abdominal wall and diaphragm were obtained at the end-exhalation phase. The combined motion vector in anterior-posterior and superior-inferior directions (AP-SI) were obtained using in-house developed software. At whole regions of pancreas, correlation coefficients (R 2 ) between pancreas motion and surrogate signals were calculated. At each pancreatic region, motion amplitude and predicted position errors for abdominal wall motion and those for diaphragmatic motion were compared. The mean positional errors between actual and predicted pancreas position based on the surrogate signal were calculated for 20 seconds.

Results The R 2 between whole pancreas motion and abdominal wall motion, diaphragm motion in AP-SI were 0.90 and 0.89, respectively. An amplitude of respiratory motion in AP-SI at head, body and tail regions of pancreas were 17.6±3.2 mm, 13.5±2.3 mm and 21.3±2.1 mm, respectively (Fig.2 (a)). The predicted position errors in AP-SI at each pancreatic region for abdominal wall motion were 1.5±0.2 mm, 1.1±0.2 mm, 2.0±0.5 mm, those for diaphragm motion were 1.4±0.3 mm, 1.2±0.1 mm, 2.1±0.6 mm, respectively (Fig.2 (b)). Although the correlation between surrogate signal and respiratory motion was high, there was a difference in motion amplitude and magnitude of predicted position error depending on the region of the pancreas.

Conclusion This study quantified the respiratory motion of each pancreatic region and characterized its relationship to