Cone-beam computed tomography (CBCT) for adaptive image guided head and neck radiation therapy.

BACKGROUND: Adaptive (ART) and image guided radiation therapy (IGRT) can improve target coverage and reduce unnecessary irradiation of organs at risk (OAR). The purpose of this study is to report the results of using mid-course imaging and a novel workflow with cone-beam computed tomography (CBCT) surveillance of dose to OAR to trigger adaptive replanning in head and neck radiotherapy. MATERIAL AND METHODS: Impact of radiation therapist (RTT) managed match protocol and mid-course imaging was assessed in two cohorts of consecutive patients receiving RT to the head and neck region, using computed tomography (CT) and CBCT-based dose verification respectively. In the CBCT cohort, patients at high risk of xerostomia received weekly dose surveillance, while low-risk patients received a mid-course CBCT review. For weekly surveillance, predicted total doses to parotid glands, spinal cord and brainstem were calculated. If predicted mean dose to parotid glands increased by >2 Gy or constraints to brainstem or spinal cord were exceeded, replanning was performed. RESULTS: None had replanning triggered by mid-course imaging. In the CBCT cohort, weekly surveillance of 40 patients yielded minimal reduction in mean dose to parotid glands of 0.65 Gy (range0.4-1 Gy ) for three patients. Patients were surveilled averagely 4.5 times during treatment. Time consumption per CBCT/week was 22 min (range 17-38). Number of patients needed to see to achieve any dose reduction to parotid glands was 13 or the equivalent of 22 working-hours. CONCLUSION: The tested dose surveillance algorithm resulted in a minimal dose reduction ( ≤1 Gy) to parotid glands for three of 40 patients. The proposed algorithm and workflow is thus not sustainable. Mid-course dose verification did not provide added benefit and can be safely omitted in the presence of closely monitored daily IGRT. Daily image guidance and match protocol is a safe and efficient method for identifying patients requiring adaptive replanning.

Accuracy of software-assisted contour propagation from planning CT to cone beam CT in head and neck radiotherapy.

BACKGROUND: Autocontouring improves workflow in computed tomography (CT)-based dose planning, but could also potentially play a role for optimal use of daily cone beam CT (CBCT) in adaptive radiotherapy. This study aims to determine the accuracy of a deformable image registration (DIR) algorithm for organs at risk (OAR) in the neck region, when applied to CBCT. MATERIAL AND METHODS: For 30 head and neck cancer (HNC) patients 14 OARs including parotid glands, swallowing structures and spinal cord were delineated. Contours were propagated by DIR from CT to the CBCTs of the first and last treatment fraction. An indirect approach, propagating contours to the first CBCT and from there to the last CBCT was also tested. Propagated contours were compared to manually corrected contours by Dice similarity coefficient (DSC) and Hausdorff distance (HD). Dose was recalculated on CBCTs and dosimetric consequences of uncertainties in DIR were reviewed. RESULTS: Mean DSC values of ≥0.8 were considered adequate and were achieved in tongue base (0.91), esophagus (0.85), glottic (0.81) and supraglottic larynx (0.83), inferior pharyngeal constrictor muscle (0.84), spinal cord (0.89) and all salivary glands in the first CBCT. For the last CBCT by direct propagation, adequate DSC values were achieved for tongue base (0.85), esophagus (0.84), spinal cord (0.87) and all salivary glands. Using indirect propagation only tongue base (0.80) and parotid glands (0.87) were ≥0.8. Mean relative dose difference between automated and corrected contours was within ±2.5% of planed dose except for esophagus inlet (-4.5%) and esophagus (5.0%) for the last CBCT using indirect propagation. CONCLUSION: Compared to manually corrected contours, the DIR algorithm was accurate for use in CBCT images of HNC patients and the minor inaccuracies had little consequence for mean dose in most clinically relevant OAR. The method can thus enable a more automated segmentation of CBCT for use in adaptive radiotherapy.

The impact of CBCT reconstruction and calibration for radiotherapy planning in the head and neck region - a phantom study.

BACKGROUND: The applicability of cone-beam computed tomography (CBCT) image sets for dose calculation purposes relies on high image quality and CT number accuracy. In this study we have investigated the use of stoichiometric calibration for transforming CT numbers into physical parameters, in combination with a new CBCT scatter correction algorithm, focusing on head and neck geometries. METHODS: CBCT projections were acquired using an On-Board-Imager (OBI v1.4; Varian Medical Systems) using both low- and high-dose clinical image acquisition protocols. The CBCT projections were reconstructed twice, using both the standard method (OBI) as well as an experimental pre-clinical reconstruction algorithm (Full Fan Experimental - FFE). Stoichiometric calibration was performed using both a phantom from CIRS with nine tissue equivalent inserts (ranging from lung to dense bone) as well as with a custom made cylindrical PMMA head and neck phantom with variable 'head' diameter and with cavities designed to fit the inserts from a Gammex RMI phantom. To benchmark the CBCT performance, the same calibration procedures were performed using two conventional CT scanners. For assessment of influence on dose-volume parameters, the head part of the anthropomorphic Alderson phantom was scanned, reconstructed with both CT and CBCT using the stoichiometric calibration curves, and finally used to compare IMRT dose calculations. RESULTS: The stoichiometric CBCT calibrations with the CIRS phantom resulted in an excellent fit between calculated and measured CT numbers (R = 1.000 for all combinations tested), equivalent to the results for the conventional scanners. Using the PMMA phantom, the stoichiometric calibration curves again showed excellent agreement, although the OBI reconstruction showed marginally increasing deviation from the unity line as the phantom size decreased. For the dose-volume parameters, deviations well within 1% were seen between the different reconstruction methods and acquisition modes. CONCLUSION: This study showed that the combination of an improved reconstruction method and stoichiometric calibration improved the CT number accuracy of CBCT scans acquired for head and neck phantoms. In particular, a radial size dependence of the scanned object similar to that in conventional CT could be achieved. Although high density inhomogeneities still are challenging for the reconstruction process, clinically acceptable agreement in key dose-volume parameters between CT-based and CBCT-based IMRT planning calculations on a humanoid phantom was found.

Normal tissue sparing in a phase II trial on daily adaptive plan selection in radiotherapy for urinary bladder cancer.

UNLABELLED: Background: Patients with urinary bladder cancer often display large changes in the shape and size of their bladder target during a course of radiotherapy (RT), making adaptive RT (ART) appealing for this tumour site. We are conducting a clinical phase II trial of daily plan selection-based ART for bladder cancer and here report dose-volume data from the first 20 patients treated in the trial. MATERIAL AND METHODS: All patients received 60 Gy in 30 fractions to the bladder; in 13 of the patients the pelvic lymph nodes were simultaneously treated to 48 Gy. Daily patient set-up was by use of cone beam computed tomography (CBCT) guidance. The first 5 fractions were delivered with large, population-based (non-adaptive) margins. The bladder contours from the CBCTs acquired in the first 4 fractions were used to create a patient-specific library of three plans, corresponding to a small, medium and large size bladder. From fraction 6, daily online plan selection was performed, where the smallest plan covering the bladder was selected prior to each treatment delivery. A total of 600 treatment fractions in the 20 patients were evaluated. RESULTS: Small, medium and large size plans were used almost equally often, with an average of 10, 9 and 11 fractions, respectively. The median volume ratio of the course-averaged PTV (PTV-ART) relative to the non-adaptive PTV was 0.70 (range: 0.46-0.89). A linear regression analysis showed a 183 cm(3) (CI 143-223 cm(3)) reduction in PTV-ART compared to the non-adaptive PTV (R(2) = 0.94). CONCLUSION: Daily adaptive plan selection in RT of bladder cancer results in a considerable normal tissue sparing, of a magnitude that we expect will translate into a clinically significant reduction of the treatment-related morbidity.

Evaluation of an application for intensity-based deformable image registration and dose accumulation in radiotherapy.

BACKGROUND: Methods to accurately accumulate doses in radiotherapy (RT) are important for tumour and normal tissues being influenced by geometric uncertainties. The purpose of this study was to investigate a pre-release deformable image registration (DIR)-based dose accumulation application, in the setting of prostate RT. MATERIAL AND METHODS: Initially accumulated bladder and prostate doses were assessed (based on 8-9 repeat CT scans/patient) for nine prostate cancer patients using an intensity-based DIR and dose accumulation algorithm as provided by the Dynamic Adaptive Radiation Therapy (DART) software. The accumulated bladder and prostate dose-volume histograms (DVHs) were compared on a range of parameters (paired Wilcoxon signed-rank test, 5% significance level) to DVHs derived using an in-house developed dose accumulation method based on biomechanical, contour-driven DIR (SurfaceRegistration). Finally, both these accumulated dose distributions were compared to the 'static' DVH, assessed from the planning CT. RESULTS: Over the population, doses accumulated with DART were overall lower than those from SurfaceRegistration (p < 0.05: D2%, gEUD and NTCP (bladder); Dmin (prostate)). The magnitude of these differences peaked for the bladder gEUD with a population median of 47 Gy for DART versus 57 Gy for SurfaceRegistration. Across the ten bladder dose/volume parameters investigated, the most pronounced individual differences were observed between the 'accumulated' DVHs and the 'static' DVHs, with deviations in mean dose up to 22 Gy. CONCLUSION: Substantial and significant differences were observed in the dose distributions between the two investigated DIR-based dose accumulation applications. The most pronounced individual differences were seen for the bladder and relative to the planned dose distribution, encouraging the use of repeat imaging data in RT planning and evaluation for this organ.

Dose/volume-based evaluation of the accuracy of deformable image registration for the rectum and bladder.

BACKGROUND AND PURPOSE: Deformable image registration (DIR) is a key component of image-guided and adaptive strategies in radiotherapy. DIR based on image intensities alone is promising for online applications, but is challenged in regions with low intensity gradients. In this study we have investigated the performance of intensity- based DIR applied to contour propagation of the rectum and bladder, focusing on the consequences in terms of dose/volume parameters. MATERIAL AND METHODS: The rectum and bladder volumes were delineated in the planning computed tomography (pCT) scan and in 8-9 repeat CTs (Vmanual) for nine prostate cancer patients. The volumes from the pCT were propagated onto the repeat CTs using intensity-based DIR (Vprop). Dose/volume parameters for Vmanual and Vprop were derived by dose re-calculations following rigid registration on prostate fiducials. Linear regression was used to identify qualitative and quantitative volumetric measures of the DIR performance being associated with the differences in dose/volume parameters. RESULTS: The median differences in dose/volume parameters assessed for Vprop and Vmanual were modest, but individual differences ~7 Gy were seen. The observed differences in dose/volume parameters showed strong correlations to the measures of the DIR performance as well as with the volume variations, most pronounced for the rectum (R(2) = 0.63-0.85; p ≤ 0.05). CONCLUSION: Limitations in the intensity-based DIR algorithm resulted in large individual differences in dose/volume parameters between propagated and manually segmented volumes, which were correlated with volumetric measures of the DIR performance.

Robust automatic segmentation of multiple implanted cylindrical gold fiducial markers in cone-beam CT projections.

PURPOSE: Implanted fiducial markers, which are used to correct for day-to-day variations, may potentially also be used to correct for intrafraction motion measurements. However, before any treatment can make use of, and react to, the position of the inserted markers they have to be segmented, either manually through expert user intervention or automatically from an imaging system. In the current study, we aimed to establish a robust and autonomous segmentation method for implanted cylindrical gold markers in a single set of projections from a cone-beam computed tomography (CBCT). METHODS: Multiple cylindrical gold markers were segmented in the projection images of CBCT scans by five sequential steps. Initially, marker candidates were identified in all projections with a blob detection routine, and then traced in subsequent projections. Traces inconsistent with a 3D marker position were rejected, and the best remaining traces were identified and used for the construction of a 3D marker constellation model, consisting of the size, position and orientation of the markers. Finally, projections of the model were used to generate templates for the final template-based marker segmentation. Hereby, challenging situations such as overlap of markers and low contrast regions were taken into account. The segmentation method was tested in 63 CBCT scans from 11 patients with 2-4 cylindrical gold markers implanted in the prostate and for 62 CBCT scans from six patients each with 2-3 cylindrical gold markers implanted in the liver and up to two cylindrical markers placed externally on the abdomen. After segmentation all projections of the 125 scans were manually inspected, and a successful segmentation was registered if the segmented position was within the projection of the marker. RESULTS: For prostate markers, the segmentation was successful in 99.8% of the projections. For the liver patients, liver markers and external markers were segmented successfully in 99.9 and 99.8% of the projections, respectively. All markers were identified in the 3D marker constellation model. The most common source of segmentation error was low contrast and motion of markers relative to each other, which resulted in a discrepancy between the template and actual projection appearance during marker overlap. Markers were overlapping in 20, 2.7, and 0.1% of the projections for prostate, liver, and external, respectively. CONCLUSIONS: We have successfully implemented a new method that, without prior knowledge on marker size, position, orientation, and number, autonomously segments cylindrical gold markers from CBCT projections with a high success rate, despite overlap, motion, and low contrast.

Evaluation of image quality for different kV cone-beam CT acquisition and reconstruction methods in the head and neck region.

PURPOSE: To evaluate the image quality obtained in a standard QA phantom with both clinical and non-clinical cone-beam computed tomography (CBCT) acquisition modes for the head and neck (HN) region as a step towards CBCT-based treatment planning. The impact of deteriorated Hounsfield unit (HU) accuracy was investigated by comparing results from clinical CBCT image reconstructions to those obtained from a pre-clinical scatter correction algorithm. METHODS: Five different CBCT acquisition modes on a clinical system for kV CBCT-guided radiotherapy were investigated. Image reconstruction was performed in both standard clinical software and with an experimental reconstruction algorithm with improved beam hardening and scatter correction. Using the Catphan 504 phantom, quantitative measures of HU uniformity, HU verification and linearity, contrast-to-noise ratio (CNR), and spatial resolution using modulation transfer function (MTF) estimation were assessed. To benchmark the CBCT image properties, comparison to standard HN protocols on conventional CT scanners was performed by similar measures. RESULTS: The HU uniformity within a water-equivalent homogeneous region was considerably improved using experimental vs. standard reconstruction, by factors of two for partial scans and four for full scans. Similarly, the amount of capping/cupping artifact was reduced by more than 1.5%. With mode and reconstruction specific HU calibration using seven inhomogeneity inserts comparable HU linearity was observed. CNR was on average 5% higher for experimental reconstruction (scaled with the square-root of dose between modes for both reconstruction methods). CONCLUSIONS: Judged on parameters affecting the common diagnostic image properties, improved beam hardening and scatter correction diminishes the difference between CBCT and CT image quality considerably. In the pursuit of CBCT-based treatment adaptation, dedicated imaging protocols may be required.

Moving metal artifact reduction in cone-beam CT scans with implanted cylindrical gold markers.

PURPOSE: Implanted gold markers for image-guided radiotherapy lead to streaking artifacts in cone-beam CT (CBCT) scans. Several methods for metal artifact reduction (MAR) have been published, but they all fail in scans with large motion. Here the authors propose and investigate a method for automatic moving metal artifact reduction (MMAR) in CBCT scans with cylindrical gold markers. METHODS: The MMAR CBCT reconstruction method has six steps. (1) Automatic segmentation of the cylindrical markers in the CBCT projections. (2) Removal of each marker in the projections by replacing the pixels within a masked area with interpolated values. (3) Reconstruction of a marker-free CBCT volume from the manipulated CBCT projections. (4) Reconstruction of a standard CBCT volume with metal artifacts from the original CBCT projections. (5) Estimation of the three-dimensional (3D) trajectory during CBCT acquisition for each marker based on the segmentation in Step 1, and identification of the smallest ellipsoidal volume that encompasses 95% of the visited 3D positions. (6) Generation of the final MMAR CBCT reconstruction from the marker-free CBCT volume of Step 3 by replacing the voxels in the 95% ellipsoid with the corresponding voxels of the standard CBCT volume of Step 4. The MMAR reconstruction was performed retrospectively using a half-fan CBCT scan for 29 consecutive stereotactic body radiation therapy patients with 2-3 gold markers implanted in the liver. The metal artifacts of the MMAR reconstructions were scored and compared with a standard MAR reconstruction by counting the streaks and by calculating the standard deviation of the Hounsfield units in a region around each marker. RESULTS: The markers were found with the same autosegmentation settings in 27 CBCT scans, while two scans needed slightly changed settings to find all markers automatically in Step 1 of the MMAR method. MMAR resulted in 15 scans with no streaking artifacts, 11 scans with 1-4 streaks, and 3 scans with severe streaking artifacts. The corresponding numbers for MAR were 8 (no streaks), 1 (1-4 streaks), and 20 (severe streaking artifacts). The MMAR method was superior to MAR in scans with more than 8 mm 3D marker motion and comparable to MAR for scans with less than 8 mm motion. In addition, the MMAR method was tested on a 4D CBCT reconstruction for which it worked equally well as for the 3D case. The markers in the 4D case had very low motion blur. CONCLUSIONS: An automatic method for MMAR in CBCT scans was proposed and shown to effectively remove almost all streaking artifacts in a large set of clinical CBCT scans with implanted gold markers in the liver. Residual streaking artifacts observed in three CBCT scans may be removed with better marker segmentation.