TH-D-AUD A-06: Study of Translational and Rotational Setup Errors and Their Correction Methods for Head & Neck Patients Using Kilovoltage Cone-Beam Computed Tomography (kV CBCT)

Abstract

Purpose: To investigate the magnitude of the six degree setup errors in head&neck patients (HNC) and evaluate which correction data, obtained from three degree and six degree 3D/3D registration, is more appropriate for setup correction if the setup errors are corrected by translational shifts. Method and Materials: kV CBCT images were acquired on the first day of treatment and weekly thereafter for 21 HNC patients treated with IMRT. A total of 145 CBCT image sets were acquired. The CBCT images were registered with the corresponding planning CT images using two different 3D rigid registration approaches. With Approach1 the registrations were conducted with translations alone, with Approach2 all six degrees were taken into account. The setup error with the maximum rotational error was simulated on planning CT for two patients, then the errors were corrected by applying the translational data obtained from Approach1 (Correction1) and Approach2 (Correction2), respectively. Dosimetric indices were compared for the two corrections. Results: For these 21 HNC patients, the average translational errors determined with Approach1 were 1.0±3.5, 0.8±3.5, 1.6±3.8mm and the values determined with Approach2 were 1.1±5.0, 0.4±3.8, 2.2±4.7mm in LR, AP and SI directions respectively. The average rotational errors determined by Approach2 were 0.6°±1.1°, 0.1°±1.9°, 0.3°±0.8° and the average maximum errors were 0.9°±1.6°, 0.5°±3.0°, 0.4°±1.1° around LR, AP and SI axes respectively. The PTV prescription dose coverage was 86.1% and 92.3% for patient1, 92.1% and 92.4% for patent2 with Correction1 and Correction2 respectively. Conclusion: Relatively large rotational errors were observed in HNC patients. Instinctively, it appeared that the Correction1 were more accurate than Correction2 if only translational corrections were involved. The result for patient1 showed that it may not be the case. The dosimetric impact of both corrective approaches has to be further investigated to evaluate which approach should be applied to correct the setup errors.