Evaluation of imaging registration errors and position setup uncertainties on image-guided radiotherapy

• Main Authors: Chia-Hao Liang, 梁家豪
• Other Authors: Jason J.S. Lee
• Format: Others
• Language: zh-TW
• Published: 2008
• Online Access: http://ndltd.ncl.edu.tw/handle/60590786047440643509

碩士 === 國立陽明大學 === 生物醫學影像暨放射科學系暨研究所 === 96 === Purpose Image-guided radiotherapy for daily patient localization and treatment guidance has become enormous popularity and clinical acceptance in recent years. Linear accelerators equipped with X-ray volumetric imagers (XVI) enable verification of location of patients and displacement of tumors in prior to treatment. Hi-Art helical tomotherapy is a highly conformal image-guided radiation therapy modality that allows for pre-treatment patient repositioning to decrease setup uncertainties. The objective of this study is to evaluate the registration accuracy using XVI system and tomotherapy to establish lower-dose imaging protocols without sacrificing registration accuracy for head routine treatment courses. Materials and methods The study was composed of two parts; one was about the XVI system and the other was for tomotherapy. In the XVI, several low-dose imaging protocols were proposed by modifying tube voltage from 120 to 100 kV and alternating tube current from 64 to 10 mA. The registration accuracies with both bone and gray value registrations provided by XVI system were performed. While in tomotherapy, modifications of slice spacing from 6 mm (coarse) to 4 mm (normal), 2 mm (fine) and scan length from 120 to 24 mm were introduced, and registration accuracies using bone and bone&tissue registration algorithms provided by Hi-Art system software was also evaluated. Results Head phantom study revealed that the gray value registration algorithm was more accurate than bone algorithm in XVI system, while with the bone registration algorithm was more superior to bone&tissue algorithm in tomotherapy. Registration accuracies between the planning CT and XVI system with protocol of 100kV and 10mA was about 1.83 ± 1.46 mm, which was reasonably acceptable in comparison with that between the planning CT and the XVI with the default protocol (p>0.05). Registration accuracies between the planning CT and tomotherapy with fixed 120mm scan length and imaging parameters of 6mm, 4mm and 2mm were 3.28 ± 0.94, 3.22 ± 0.28, 2.11 ± 0.51mm, respectively. In addition, those between planning CT and tomotherapy with imaging parameters of fixed 24mm scan length and variation of slice spacing (6, 4mm); fixed 30mm scan length and variation of slice spacing (6, 4mm) was 3.29 ± 0.63, 3.27 ± 0.57, 3.36 ± 0.42 and 3.28 ± 0.71mm, respectively. The results revealed that registration accuracies between the planning CT and tomotherapy using bone registration method with imaging parameters of slice spacing (6, 4mm) and different scan length (120, 30, 24mm) were similar. Conclusion Image-guided radiotherapies are state-of-the-art treatment systems that could deliver highly conformal dose distribution with very high degrees of accuracy in geometry and dosimetry. Phantom setup errors between the reference planning CT and XVI or tomotherapy were studied for compromise purpose between radiation dose and registration accuracy. In conclusion, the imaging protocol in the XVI system with 100kV and 10ms, while that in the Hi-Art tomotherapy with 6mm slice spacing and 24-30mm scan length seemed to be optimal for patient localization purpose. The suggested lower-dose protocols would result in less radiation dose to patients in positioning and yet without compromising registration accuracy in both systems, and may have the potential to be adopted for clinical usage in the future. Keywords: image-guided radiotherapy, tomotherapy, image registration, registration accuracy