Estimation of lung tumor position from multiple anatomical features on 4D-CT using multiple regression analysis.

To estimate the lung tumor position from multiple anatomical features on four-dimensional computed tomography (4D-CT) data sets using single regression analysis (SRA) and multiple regression analysis (MRA) approach and evaluate an impact of the approach on internal target volume (ITV) for stereotactic body radiotherapy (SBRT) of the lung. Eleven consecutive lung cancer patients (12 cases) underwent 4D-CT scanning. The three-dimensional (3D) lung tumor motion exceeded 5 mm. The 3D tumor position and anatomical features, including lung volume, diaphragm, abdominal wall, and chest wall positions, were measured on 4D-CT images. The tumor position was estimated by SRA using each anatomical feature and MRA using all anatomical features. The difference between the actual and estimated tumor positions was defined as the root-mean-square error (RMSE). A standard partial regression coefficient for the MRA was evaluated. The 3D lung tumor position showed a high correlation with the lung volume (R = 0.92 ± 0.10). Additionally, ITVs derived from SRA and MRA approaches were compared with ITV derived from contouring gross tumor volumes on all 10 phases of the 4D-CT (conventional ITV). The RMSE of the SRA was within 3.7 mm in all directions. Also, the RMSE of the MRA was within 1.6 mm in all directions. The standard partial regression coefficient for the lung volume was the largest and had the most influence on the estimated tumor position. Compared with conventional ITV, average percentage decrease of ITV were 31.9% and 38.3% using SRA and MRA approaches, respectively. The estimation accuracy of lung tumor position was improved by the MRA approach, which provided smaller ITV than conventional ITV.

Difference in dose-volumetric data between the analytical anisotropic algorithm, the dose-to-medium, and the dose-to-water reporting modes of the Acuros XB for lung stereotactic body radiation therapy.

The purpose of this study was to evaluate the difference in dose-volumetric data between the analytical anisotropic algorithms (AAA) and the two dose reporting modes of the Acuros XB, namely, the dose to water (AXB_Dw) and dose to medium (AXB_Dm) in lung stereotactic body radiotherapy (SBRT). Thirty-eight plans were generated using the AXB_Dm in Eclipse Treatment Planning System (TPS) and then recalculated with the AXB_Dw and AAA, using identical beam setup. A dose of 50 Gy in 4 fractions was prescribed to the isocenter and the planning target volume (PTV) D95%. The isocenter was always inside the PTV. The following dose-volumetric parameters were evaluated; D2%, D50%, D95%, and D98% for the internal target volume (ITV) and the PTV. Two-tailed paired Student's t-tests determined the statistical significance. Although for most of the parameters evaluated, the mean differences observed between the AAA, AXB_Dm, and AXB_Dw were statistically significant (p < 0.05), absolute differences were rather small, in general less than 5% points. The maximum mean difference was observed in the ITV D50% between the AXB_Dm and the AAA and was 1.7% points under the isocenter prescription and 3.3% points under the D95 prescription. AXB_Dm produced higher values than AXB_Dw with differences ranging from 0.4 to 1.1% points under isocenter prescription and 0.0 to 0.7% points under the PTV D95% prescription. The differences observed under the PTV D95% prescription were larger compared to those observed for the isocenter prescription between AXB_Dm and AAA, AXB_Dm and AXB_Dw, and AXB_Dw and AAA. Although statistically significant, the mean differences between the three algorithms are within 3.3% points.

Evaluation of different set-up error corrections on dose-volume metrics in prostate IMRT using CBCT images.

We investigated the effect of different set-up error corrections on dose-volume metrics in intensity-modulated radiotherapy (IMRT) for prostate cancer under different planning target volume (PTV) margin settings using cone-beam computed tomography (CBCT) images. A total of 30 consecutive patients who underwent IMRT for prostate cancer were retrospectively analysed, and 7-14 CBCT datasets were acquired per patient. Interfractional variations in dose-volume metrics were evaluated under six different set-up error corrections, including tattoo, bony anatomy, and four different target matching groups. Set-up errors were incorporated into planning the isocenter position, and dose distributions were recalculated on CBCT images. These processes were repeated under two different PTV margin settings. In the on-line bony anatomy matching groups, systematic error (∑) was 0.3 mm, 1.4 mm, and 0.3 mm in the left-right, anterior-posterior (AP), and superior-inferior directions, respectively. ∑ in three successive off-line target matchings was finally comparable with that in the on-line bony anatomy matching in the AP direction. Although doses to the rectum and bladder wall were reduced for a small PTV margin, averaged reductions in the volume receiving 100% of the prescription dose from planning were within 2.5% under all PTV margin settings for all correction groups, with the exception of the tattoo set-up error correction only (≥ 5.0%). Analysis of variance showed no significant difference between on-line bony anatomy matching and target matching. While variations between the planned and delivered doses were smallest when target matching was applied, the use of bony anatomy matching still ensured the planned doses.

Multivariate analysis of factors predicting prostate dose in intensity-modulated radiotherapy.

We conducted a multivariate analysis to determine relationships between prostate radiation dose and the state of surrounding organs, including organ volumes and the internal angle of the levator ani muscle (LAM), based on cone-beam computed tomography (CBCT) images after bone matching. We analyzed 270 CBCT data sets from 30 consecutive patients receiving intensity-modulated radiation therapy for prostate cancer. With patients in the supine position on a couch with the HipFix system, data for center of mass (COM) displacement of the prostate and the state of individual organs were acquired and compared between planning CT and CBCT scans. Dose distributions were then recalculated based on CBCT images. The relative effects of factors on the variance in COM, dose covering 95% of the prostate volume (D95%), and percentage of prostate volume covered by the 100% isodose line (V100%) were evaluated by a backward stepwise multiple regression analysis. COM displacement in the anterior-posterior direction (COMAP) correlated significantly with the rectum volume (δVr) and the internal LAM angle (δθ; R = 0.63). Weak correlations were seen for COM in the left-right (R = 0.18) and superior-inferior directions (R = 0.31). Strong correlations between COMAP and prostate D95% and V100% were observed (R ≥ 0.69). Additionally, the change ratios in δVr and δθ remained as predictors of prostate D95% and V100%. This study shows statistically that maintaining the same rectum volume and LAM state for both the planning CT simulation and treatment is important to ensure the correct prostate dose in the supine position with bone matching.