Deformable Image Registration Uncertainty Research Articles

SU‐E‐J‐62: Estimating Plausible Treatment Course Dose Distributions by Accounting for Registration Uncertainty and Organ Motion

Purpose:Dose accumulation following deformable image registration (DIR) is challenging. In this study, we used a statistical sampling approach, which takes into account both DIR uncertainties and patient‐specific organ motion, to study the distribution of possible true dose distributions.Methods:The study included ten patients (six CT scans/patient) treated with radiotherapy for prostate cancer. For each patient, the planned dose was re‐calculated on the repeated geometries, following rigid registration based on fiducial markers. The dose re‐calculated on the first CT served as our snapshot dose distribution (D1) and the average of the first five repeat scans as our treatment course reference dose distribution (Dref). Patient‐specific motion and DIR‐uncertainties, at each voxel in CT1, were assessed using a previously developed DIR performance measure, the distance discordance metric (DDM). To sample the distribution of possible true, predicted dose distributions (Dpred), we resampled D1 by perturbing the location of each voxel with the corresponding DDM. The three dose distribution approaches are compared for the rectum and the bladder.Results:The bladder generalized equivalent uniform dose (gEUD) from the averaged Dpred was closer to the gEUDref than to the gEUD1 (difference: 0.6 vs. 1.0 Gy). For both structures, the gEUDpred was higher than the gEUDref, and significantly higher (p≤0.05) for the rectum (average: 50.8 Gy vs. 48.0 Gy).Conclusion:We have shown that the bladder gEUD values resulting from our DIR‐uncertainty inclusive dose sampling approach, Dpred, were closer to the gEUD from Dref than the gEUD values from D1. For the rectum, gEUDpred overestimated gEUDref. Theoretically however, gEUDpred values, sampled from DDM uncertainties are more representative of dose uncertainties.

Voxel-based statistical analysis of uncertainties associated with deformable image registration

Deformable image registration (DIR) algorithms have inherent uncertainties in their displacement vector fields (DVFs).The purpose of this study is to develop an optimal metric to estimate DIR uncertainties. Six computational phantoms have been developed from the CT images of lung cancer patients using a finite element method (FEM). The FEM generated DVFs were used as a standard for registrations performed on each of these phantoms. A mechanics-based metric, unbalanced energy (UE), was developed to evaluate these registration DVFs. The potential correlation between UE and DIR errors was explored using multivariate analysis, and the results were validated by landmark approach and compared with two other error metrics: DVF inverse consistency (IC) and image intensity difference (ID). Landmark-based validation was performed using the POPI-model. The results show that the Pearson correlation coefficient between UE and DIR error is rUE-error = 0.50. This is higher than rIC-error = 0.29 for IC and DIR error and rID-error = 0.37 for ID and DIR error. The Pearson correlation coefficient between UE and the product of the DIR displacements and errors is rUE-error × DVF = 0.62 for the six patients and rUE-error × DVF = 0.73 for the POPI-model data. It has been demonstrated that UE has a strong correlation with DIR errors, and the UE metric outperforms the IC and ID metrics in estimating DIR uncertainties. The quantified UE metric can be a useful tool for adaptive treatment strategies, including probability-based adaptive treatment planning.

Dose mapping sensitivity to deformable registration uncertainties in fractionated radiotherapy - applied to prostate proton treatments.

BackgroundCalculation of accumulated dose in fractionated radiotherapy based on spatial mapping of the dose points generally requires deformable image registration (DIR). The accuracy of the accumulated dose thus depends heavily on the DIR quality. This motivates investigations of how the registration uncertainty influences dose planning objectives and treatment outcome predictions.A framework was developed where the dose mapping can be associated with a variable known uncertainty to simulate the DIR uncertainties in a clinical workflow. The framework enabled us to study the dependence of dose planning metrics, and the predicted treatment outcome, on the DIR uncertainty. The additional planning margin needed to compensate for the dose mapping uncertainties can also be determined. We applied the simulation framework to a hypofractionated proton treatment of the prostate using two different scanning beam spot sizes to also study the dose mapping sensitivity to penumbra widths.ResultsThe planning parameter most sensitive to the DIR uncertainty was found to be the target D95. We found that the registration mean absolute error needs to be ≤0.20 cm to obtain an uncertainty better than 3% of the calculated D95 for intermediate sized penumbras. Use of larger margins in constructing PTV from CTV relaxed the registration uncertainty requirements to the cost of increased dose burdens to the surrounding organs at risk.ConclusionsThe DIR uncertainty requirements should be considered in an adaptive radiotherapy workflow since this uncertainty can have significant impact on the accumulated dose. The simulation framework enabled quantification of the accuracy requirement for DIR algorithms to provide satisfactory clinical accuracy in the accumulated dose.

SU‐E‐J‐83: Evaluation of Landmark‐Assisted Deformable Image Registration Technique On Inter‐and Intra‐Patient Head and Neck CT Imaging

Purpose: Accurate image registration in patients with Head and Neck (HN) cancer is critical for assessment of treatment response, and must be rigorously validated. To evaluate deformable image registration (DIR) with our previously developed DIR techniques on HN PET‐CT scans, we incorporated distance discordance metric (DDM), a novel method for quantification of local registration uncertainty, as well as traditional quantitative and qualitative measurements. Methods: For each CT image series, six landmark (LM) points relevant for patients with HN cancer were placed, including the inferior‐most extent of the styloid processes, the lateral‐and inferior‐most aspects of the hyoid bone, and the superior central face of the sternum. These points were used either to assist or assess registration.Inter‐patient HN CT image registration for 10 patients was performed using standard intensity‐based B‐Spline and LM‐assisted B‐Spline DIR techniques. Intra‐patient DIR of the same patients was performed using pre‐and post‐radiation therapy image sets of the same patient.DDM metrics were obtained for inter‐patient DIR results and compared to traditional qualitative and quantitative methods including whole image mean square error (MSE), relative landmark distances between scans, and visual judgment by physicians. Results: Registration results were improved with the addition of landmarks on bony structures relevant to HN cancer. While MSE only provides a general estimate of the overall image registration error, the novel DDM metric shows the registration uncertainties for each pixel on the fixed image, allowing for localization of error. Based on DDM, DIR uncertainty was significantly reduced (>50% in region of interest) by LM‐assisted DIR techniques on inter‐patient registration. Relative LM distance was reduced as well for both intra‐and inter‐patient DIR (<1mm). Conclusion: Evaluation of DIR results with novel and traditional methods demonstrates the LM‐assisted DIR approach achieves better registration results on both inter‐and intra‐patient DIR.

TU‐C‐141‐09: An Automated Workflow for Patient‐Specific Verification of Deformable Image Registration

Purpose: To develop an automated workflow for patient‐specific verification of deformable image registration (DIR). Methods: The central concept behind this verification process is digitally applying test deformations to patient images and comparing to those returned from clinical DIR algorithms. The patient‐specific nature of these test deformations is the novel feature of this workflow. First, a set of different test DIR algorithms is applied to the original clinical images. The deformations produced by these algorithms represent plausible deformations and are applied to the clinical moving image, producing a set of test fixed images. A combination of computed‐tomography images of a water phantom with frequency‐domain image filtering is utilized to first remove image noise and then reintroduce distinct noise fluctuations on the moving and fixed test images, which would otherwise skew DIR error evaluation. The clinical algorithm is then applied to the set of test images, which yields a set of theoretical deformations. For each image voxel, the set of theoretical deformations are compared to those actually applied. The maximum discrepancy in each direction becomes the theoretical uncertainty for that voxel on the clinical image. This creates an uncertainty vector field, which can be applied to patient contours and dose to evaluate potential clinical errors from DIR uncertainty. The results of this process are benchmarked against an anthropomorphic pelvic phantom, which possesses a measured ground‐truth deformation. Results: The theoretical DIR uncertainties from the workflow gave an accurate indication for the spatial location of the actual errors, but tended to overestimate the error magnitude. The maximum theoretical uncertainties in the anterior and lateral directions were 15.4 and 11.5 mm, respectively, compared to the maximum measured errors of 12.6 and 3.0 mm, respectively. Conclusion: This workflow demonstrates the ability to perform automated patient‐specific verification for DIR. Further work to develop the method is ongoing.

A novel four-dimensional radiotherapy planning strategy from a tumor-tracking beam's eye view

To investigate the feasibility of four-dimensional radiotherapy (4DRT) planning from a tumor-tracking beam's eye view (ttBEV) with reliable gross tumor volume (GTV) delineation, realistic normal tissue representation, high planning accuracy and low clinical workload, we propose and validate a novel 4D conformal planning strategy based on a synthesized 3.5D computed tomographic (3.5DCT) image with a motion-compensated tumor. To recreate patient anatomy from a ttBEV in the moving tumor coordinate system for 4DRT planning (or 4D planning), the centers of delineated GTVs in all phase CT images of 4DCT were aligned, and then the aligned CTs were averaged to produce a new 3.5DCT image. This GTV-motion-compensated CT contains a motionless target (with motion artifacts minimized) and motion-blurred normal tissues (with a realistic temporal density average). Semi-automatic threshold-based segmentation of the tumor, lung and body was applied, while manual delineation was used for other organs at risk (OARs). To validate this 3.5DCT-based 4D planning strategy, five patients with peripheral lung lesions of small size (<5 cm3) and large motion range (1.2–3.5 cm) were retrospectively studied for stereotactic body radiotherapy (SBRT) using 3D conformal radiotherapy planning tools. The 3.5DCT-based 4D plan (3.5DCT plan) with 9–10 conformal beams was compared with the 4DCT-based 4D plan (4DCT plan). The 4DCT plan was derived from multiple 3D plans based on all phase CT images, each of which used the same conformal beam configuration but with an isocenter shift to aim at the moving tumor and a minor beam aperture and weighting adjustment to maintain plan conformality. The dose–volume histogram (DVH) of the 4DCT plan was created with two methods: one is an integrated DVH (iDVH4D), which is defined as the temporal average of all 3D-phase-plan DVHs, and the other (DVH4D) is based on the dose distribution in a reference phase CT image by dose warping from all phase plans using the displacement vector field (DVF) from a free-form deformable image registration (DIR). The DVH3.5D (for the 3.5DCT plan) was compared with both iDVH4D and DVH4D. To quantify the DVH difference between the 3.5DCT plan and the 4DCT plan, two methods were used: relative difference (%) of the areas underneath the DVH curves and the volumes receiving more than 20% (V20) and 50% (V50) of prescribed dose of these 4D plans. The volume of the delineated GTV from different phase CTs varied dramatically from 24% to 112% among the five patients, whereas the GTV from 3.5DCT deviated from the averaged GTV in 4DCT by only −6%±6%. For planning tumor volume (PTV) coverage, the difference between the DVH3.5D and iDVH4D was negligible (<1% area), whereas the DVH3.5D and DVH4D were quite different, due to DIR uncertainty (∼2 mm), which propagates to PTV dose coverage with a pronounced uncertainty for small tumors (0.3–4.0 cm3) in stereotactic plans with sharp dose falloff around PTV. For OARs, such as the lung, heart, cord and esophagus, the three DVH curves (DVH3.5D, DVH4D and iDVH4D) were found to be almost identical for the same patients, especially in high-dose regions. For the tumor-containing lung, the relative difference of the areas underneath the DVH curves was found to be small (5.3% area on average), of which 65% resulted from the low-dose region (D < 20%). The averaged V20 difference between the two 4D plans was 1.2% ± 0.8%. For the mean lung dose (MLD), the 3.5DCT plan differed from the 4DCT plan by −1.1%±1.3%. GTV-motion-compensated CT (3.5DCT) produces an accurate and reliable GTV delineation, which is close to the mean GTV from 4DCT. The 3.5DCT plan is equivalent to the 4DCT plan with <1% dose difference to the PTV and negligible dose difference in OARs. The 3.5DCT approach simplifies 4D planning and provides accurate dose calculation without a substantial increase of clinical workload for motion-tracking delivery to treat small peripheral lung tumors with large motion.

SU-E-J-120: Characterization of 4D Lung Ventilation Based on 4DCT Lung Density.

To characterize patient's 4D lung ventilation and motion using 4DCT images aiming to extract pulmonary features for developing a biomechanical model to predict lung tumor motion. 4DCT images of sixteen patients were used for initial 4D-ventilation evaluation. A program was developed for visualization and quantification of 4D lung ventilation. Lung density gradient and pleural pressure variation were evaluated based on the gravity. Freeform deformable image registration (DIR) was applied to map the lung density from one phase to another, with linearly interpolated voxel density. The warped lung density from a secondary CT was compared with the primary CT (full-exhalation) for quantification. To minimize the uncertainty from motion artifacts for quantitative ventilation/motion analysis, we developed two screening algorithms (based on Fourier Transformation and Generalized Linear Regression) to quantify the regularity (score: 0 to 1) of the RPM waveform, assuming the higher the regularity in RPM the less motion artifacts in 4DCT. Ninety-seven RPM curves (54 free-breathing and 43 audio-coaching patients) were used to qualitatively test these screening methods. Lung density gradient in anterior-posterior direction for supine patients is observed in 15/16 patients at scale of >100-200HU. The gravity pressure due to lung tissue weight is estimated to be ∼4mmHg, which could neutralize the pleural pressure at the posterior chest wall. The different pleural pressures at anterior/posterior chest walls may explain the difference in local diaphragm muscle engagement and motion range: 8±5mm (anterior) and 23±12mm (posterior). Motion artifacts and DIR uncertainty affected quantitative ventilation mapping. In the RPM regularity assessment, audiocoaching patients score higher than free-breathing patients: 0.76±0.13 vs. 0.50±0.19 (Fourier method) and 0.74±0.19 vs. 0.66±0.17 (Regression method). High-quality 4DCTs can be readily selected for further quantitative ventilation/motion analysis. We have characterized several basic features of lung ventilation and motion and developed two effective screening methods for high-quality 4DCT for on-going quantitative studies.

SU‐E‐J‐121: A New 4D Radiotherapy Planning Strategy Using Synthesized Tumor‐Motion‐Compensated Computed Tomography

To validate a four-dimensional radiotherapy (4DRT) planning based on a synthesized CT image compensating for tumor motion, accounting for tumor rotation, deformation and distortion due to motion artifacts, and producing realistic normal tissue density in motion-tracking beam eye view. 4D computed tomography (4DCT) images of six patients with peripheral lung lesions in mid or lower lobs (motion range: 0.5-3.5cm and size: 1.5±2.0 cm3 ) were used. A customized program was used for simulating the patient anatomy with a motion-compensated tumor using 4DCT by aligning the tumor and averaging the 4DCT into a static 3.5DCT image. The gross tumor volume (GTV) was delineated semiautomatically using a threshold algorithm. Variation of GTV in each phase CT was assessed across all phases. 3DRT plans were generated using 3.5DCT and 4DCT and compared for validation. An integrated dose volume histogram (iDVH) from all phase plans and dose warping using deformable image registration (DIR) were used for evaluating 4D plans and comparing with 3.5D plans. The range of tumor volume variation over the mean within a breathing cycle was 87%±46%. The 3.5DCT produced an 'averaged' GTV, more reliable than that from any phase CT. The results show that the 3.5D plan is equivalent to the 4D plan, except for low dose area, using the iDVH evaluation. On average, the percentage difference for the areas under the DVH and iDVH is 4.3%±2.7%, while 2/3 of the difference results from low dose region blow D20%. Using DIR-based dose warping, PTV coverage varies due to DIR uncertainty for the small lesions. The 3.5D plan is equivalent to the 4D plan for peripheral lung lesions, yet requires much less clinical workload. The 3.5D plan accounts for tumor motion and tumor variation for a more reliable delineation, and for realistic normal tissue representation for motion tracking.

Estimation of three‐dimensional intrinsic dosimetric uncertainties resulting from using deformable image registration for dose mapping

This article presents a general procedural framework to assess the point-by-point precision in mapped dose associated with the intrinsic uncertainty of a deformable image registration (DIR) for any arbitrary patient. Dose uncertainty is obtained via a three-step process. In the first step, for each voxel in an imaging pair, a cluster of points is obtained by an iterative DIR procedure. In the second step, the dispersion of the points due to the imprecision of the DIR method is used to compute the spatial uncertainty. Two different ways to quantify the spatial uncertainty are presented in this work. Method A consists of a one-dimensional analysis of the modules of the position vectors, whereas method B performs a more detailed 3D analysis of the coordinates of the points. In the third step, the resulting spatial uncertainty estimates are used in combination with the mapped dose distribution to compute the point-by-point dose standard deviation. The process is demonstrated to estimate the dose uncertainty induced by mapping a 62.6 Gy dose delivered on maximum exhale to maximum inhale of a ten-phase four-dimensional lung CT. For the demonstration lung image pair, the standard deviation of inconsistency vectors is found to be up to 9.2 mm with a mean sigma of 1.3 mm. This uncertainty results in a maximum estimated dose uncertainty of 29.65 Gy if method A is used and 21.81 Gy for method B. The calculated volume with dose uncertainty above 10.00 Gy is 602 cm3 for method A and 1422 cm3 for method B. This procedure represents a useful tool to evaluate the precision of a mapped dose distribution due to the intrinsic DIR uncertainty in a patient. The procedure is flexible, allowing incorporation of alternative intrinsic error models.

SU‐GG‐J‐45: Application of Virtual Patient Model in 4D IMRT Monte Carlo Treatment Planning

Purpose: 4D IMRT planning using 4D CT images has uncertainties such as image artifacts, image sorting uncertainty and deformable image registration uncertainty. On the other hand, using virtual reality to create patient‐specific virtual model can represent the patient more accurately. Therefore, in this study, we are aiming to investigate the feasibility of 4D IMRT planning using virtual patient model. Method and Materials: The input contours for virtual model was from Pinnacle treatment planning system. The polygon of a structure formed from sampled control points, were used to reconstruct the NURBS surface of this structure. A 4D virtual patient was initially a sequence of 3D NURBS models at different phases. To be applied in real‐time breathing management, a few methods were used to adjust the model to account for the variation of breathing and to give model the ability to predict motion. Kaiman filter was used to predict the breathing curve. Two Monte Carlo codes, MCSIM and EGS4‐VLSI were used to simulate the treatment plan using the opening density matrix exported from Pinnacle optimization results. MCsim used patient's CT in dose calculation while EGS4‐VLSI used a phantom converted from patient's model. DVHS were compared for end‐exhale phase among the results from Pinnacle and the two Monte Carlo codes. Results: Monte Carlo codes benchmark showed good agreement for clinical measured data for 10×10 cm2 PDD and profiles. DVHs showed differences mainly due to the inhomogeneity effects for dose calculation. Conclusion: A virtual model based 4D IMRT method was proposed and preliminary results show the feasibility of this methodology. Varity exits among different calculation method, real clinical benefit of model based planning needs further investigation.Conflict of interest: Research supported by grant R0I LM009362.