ABSTRACT
Objective To analyze setup errors in cone beam computed tomography(CBCT)-guided radiotherapy for nasopharyngeal cancer(NPC)and evaluate the use frequency of CBCT. Methods Twenty-six patients newly diagnosed with NPC who received CBCT scan no less than 3 times per week during the whole course(7 weeks)of intensity-modulated radiotherapy(IMRT)were enrolled as subjects. Two setup errors were recorded each week:the setup error in the first CBCT analysis(Ef)and the setup error in the subsequent CBCT analysis(Ec). Comparisons of Efbetween two different weeks, Ecbetween two different weeks,and between Efand Ecin the same week were made by analysis of variance. Results During the 7 weeks,there were no significant differences in Efin the medial-lateral(ML), superior-inferior(SI), or anterior-posterior(AP)direction between any two weeks(all P>0.05);there were also no significant differences in Ecin the ML or SI direction between any two weeks(all P>0.05);there was a significant difference in Ecin the AP direction between two different weeks(P<0.05);there were no significant differences between Efand Ecin the ML or SI direction in any week(P>0.05);there were also no significant differences between Efand Ecin the AP direction in any week(P>0.05)except for the first week (P<0.05). Conclusions In IMRT for NPC,the setup errors in the ML and SI directions are stable in the whole course of radiotherapy,while the setup errors in the AP direction are different between the first week and subsequent weeks. Daily CBCT scan in the first week and weekly CBCT scan in the subsequent weeks are highly recommended for most patients to reduce the use frequency of CBCT.
ABSTRACT
Objective To compare the results of three different registration methods in the kilovolt CBCT guided IMRT for nasopharyngeal cancer (NPC).Methods Total 560 CBCT images of 24 NPC patients who received kilovolt CBCT guided IMRT were analyzed off line.Three registration methods were used for alignment between CBCT and planning CT,including translational and rotational errors for bone and grey (BoneT + R,GreyT+ R),only translational errors for grey (GreyT).The registration results were analyzed by mean paired t-test respectively.Results With method BoneT+R,the translational errors on x,y and z axes were (-0.11 ± 1.35) mm,(0.40 ± 2.09) mm and (0.95 ± 1.56) mm and the rotational errors were 1.06° ±0.67°,0.01° ± 1.28° and 0.92° ± 1.00° respectively.With GreyT+R,the translational errors on x,y and z axes were (-0.02 ± 1.06) mm,(0.68 ± 1.92) mm and (0.81 ± 1.46) mm and the rotational errors were 0.85° ±0.61°,-0.05° ± 1.32° and 0.91° ±0.72° respectively.With GreyT,the translational errors on x,y and z axes were (0.58 ± 1.02),(0.52 ± 1.89) and (0.44 ± 1.43) mm.The results of compared mean t-test for different registration methods groups have significant difference (P =0.00-0.01) except for the rotational errors on y and z axes between BoneT+R and GreyT+R (P =0.05,0.62).Conclusions There have different alignment errors when different registration methods used for NPC kilovolt CBCT guided radiotherapy.If there have correct methods for rotation errors,GreyT+R registration method may be a better choice.In opposite,GreyT+R registration method would be used firstly to verify whether the rotational error ＞ 2°or 3°.If the rotational error ＞ 2°or 3°,the patient should be re-setup.If not,according to these alignment results,the GreyT method,manual method would be used to compensate the translational errors.