Development of a Low-Cost and Clinically Available Patient Intrafraction Motion Monitoring System
Y Uchida1*, H Tachibana2, Y Takada1, K Kashihara1, T Yamashita1, (1) Tokyo Radiation Oncology Clinic Koutou,Tokyo,Japan,(2) UT Southwestern Medical Center, DALLAS, TX,USASU-E-J-165 Sunday 3:00:00 PM - 6:00:00 PM Room: Exhibit Hall
Purpose: We have developed a flexible system in order to monitor the intrafraction motion. The purpose of our study is to extend the monitoring system to clinically available system with low cost.
Methods: Our system is composed of a standard web camera with a low-cost and 8x magnitude telescope, a personal computer with our in-house software and a specific marker box. The telescopic lens was attached with the web camera to extend more effective distance of the measurement in order to avoid patient's interference. The dynamic calibration algorithm was developed to take into account patient's rotation during treatment in order to measure the intrafraction motion more accurately. Tracking three markers simultaneously based on a template matching technique using parallel CPU computing was performed to measure the intrafraction motion with dynamic calibration. To evaluate our new system, a respiratory motion QA phantom with 10 mm-amplitude was used in order to measure the amplitudes under the different angles of web camera setting (0 to 50 degree, 5 degree step) using our system and Varian Real-Time Position Management Respiratory Gating (Varian-RPM) System. The results of our system were compared with the results of Varian-RPM System.
Results: The result of the amplitudes measured by our system and Varian RPM-System are 10.2±0.3 mm and 10.3±0.1mm at the angle of 0 deg., respectively. The values of both systems were within the tolerance value of AAPM Task group 142. The results of the amplitude of our system and Varian-RPM system were 10.2±0.3mm and 10.4±0.1mm, respectively, while the angle was changed. Under the parallel CPU computing, the calculation time to measure the position of the marker was about 50msec including the latency.
Conclusions: Our proposed system could have clinically acceptable accuracies. The system would be contributed broadly to improve the treatment accuracy because of low-cost installation of it.