Impact Of Geometric Uncertainties Research Articles

Assessment of robustness of institutional applied clinical target volume (CTV) to planning target volume (PTV) margin in cervical cancer using biological models

The aim of this study is to investigate the robustness of our institutionally applied clinical target volume (CTV)-to-planning target volume (PTV) margins in cervical cancer patients in terms of an equivalent uniform dose (EUD) based on tumor control probability (TCP). We simulated target motion using 25 IMRT cervical cancer plans to demonstrate the effect of geometrical uncertainties on the EUD and TCP. The different components of the total geometrical uncertainties budget were estimated. The biological effects were compared by calculating the EUDs from the trial DVHs. The impact of geometric uncertainties was calculated as a percentage of the difference between 〖EUD〗_static and 〖EUD〗_motion, where the 〖EUD〗_static is the EUD calculated from the target DVHs and 〖EUD〗_motion is averaged, over a 1000 calculated EUDs for each of the analyzed IMRT treatment plans. The multivariate nonlinear regression was used to find the predicted difference between the static and motion EUD. The estimate of the systematic and random motion errors were Σ_(total(SI,LR,AP)) (mm)=(2.6; 2.5; 1.8) and σ_(total(SI,LR,AP)) (mm)=(3.4; 1.4; 3.4). For average 〈EUD〉_motion=44.3 Gy (over 25 patients) we have found a TCP decrease of about 1%, %(ΔTCP)≈1% for predefined PTV margin. According to the calculated EUD motion-distributions, for particular patients, the CTV does receive the prescribed EUD of 45 Gy. The predicted difference in EUD showed that our isotropic margin of 10 mm is large enough to absorb geometric uncertainties and ensure dose coverage of the moving CTV in the cervical cancer patients.

Effects of left ventricle wall thickness uncertainties on cardiac mechanics.

Computational models of the heart have reached a level of maturity that enables sophisticated patient-specific simulations and hold potential for important applications in diagnosis and therapy planning. However, such clinical use puts strict demands on the reliability and accuracy of the models and requires the sensitivity of the model predictions due to errors and uncertainty in the model inputs to be quantified. The models typically contain a large number of parameters, which are difficult to measure and therefore associated with considerable uncertainty. Additionally, patient-specific geometries are usually constructed by semi-manual processing of medical images and must be assumed to be a potential source of model uncertainty. In this paper, we assess the model accuracy by considering the impact of geometrical uncertainties, which typically occur in image-based computational geometries. An approach based on 17 AHA segments diagram is used to consider uncertainties in wall thickness and also in the material properties and fiber orientation, and we perform a comprehensive uncertainty quantification and sensitivity analysis based on polynomial chaos expansions. The quantities considered include stress, strain and global deformation parameters of the left ventricle. The results indicate that important quantities of interest may be more affected by wall thickness, and highlight the need for accurate geometry reconstructions in patient-specific cardiac mechanics models.

Impact of geometric uncertainty on hemodynamic simulations using machine learning

Abstract In the cardiovascular system, blood flow rates, blood velocities and blood pressures can be modeled using the Navier–Stokes equations. Inputs to the system are typically uncertain, such as (a) the geometry of the arterial tree, (b) clinically measured blood pressure and viscosity, (c) boundary resistances, among others. Due to a large number of such parameters, efficient quantification of uncertainty in solution fields in this multi-parameter space is challenging. We use an adaptive stochastic collocation method to quantify the impact of uncertainty in geometry in patient-specific models. We develop a novel subdivision method to define the stochastic space of geometries. To accelerate convergence and make the problem tractable, we use a machine learning approach to approximate the simulation-based solution. Towards this, a reduced order model of the Navier–Stokes equations is developed using a segmental resistance analog boundary conditions (ratio of pressure to flow). Using an offline database of pre-computed solutions, we compute a map (rule) from the features to solution fields. We achieve significant speed-up (of a few orders of magnitude) by approximating the simulation-based solution using a machine learning predictor. A bootstrap aggregated decision tree was found to be the best predictor among many candidate regressors (correlation coefficient of training set was 0.94). We demonstrate stochastic space convergence using the adaptive stochastic collocation method, and also show robustness to the choice of geometry parameterization. The sensitivities to geometry obtained using machine learning had a correlation coefficient of 0.92 with the values obtained using finite element simulations. Segments with significant disease in the larger arteries had the highest sensitivities. Terminal segments are more sensitive to dilation and proximal healthy segments are more sensitive to erosion. Sensitivity to geometry is highest when geometric resistance is comparable to net downstream resistance.

SP-0504: Impact of geometrical uncertainties in extreme hypo with brachytherapy

SP-0504: Impact of geometrical uncertainties in extreme hypo with brachytherapy

SP-0503: Impact of geometrical uncertainties in stereotactic radiation therapy: risk assessment and clinical management

SP-0503: Impact of geometrical uncertainties in stereotactic radiation therapy: risk assessment and clinical management

The missing complexity in seismically imaged normal faults: what are the implications for geometry and production response?

Abstract The impact of geometric uncertainty on across-fault flow behaviour at the scale of individual intra-reservoir faults is investigated in this study. A high resolution digital elevation model (DEM) of a faulted outcrop is used to construct an outcrop-scale geocellular grid capturing high-resolution fault geometries (5 m scale). Seismic forward modelling of this grid allows generation of a 3D synthetic seismic cube, which reveals the corresponding seismically resolvable fault geometries (12.5 m scale). Construction of a second geocellular model, based upon the seismically resolvable fault geometries, allows comparison with the original outcrop geometries. Running fluid flow simulations across both models enables us to assess quantitatively the impact of outcrop resolution v. seismic resolution fault geometries upon across-fault flow. The results suggest that seismically resolvable fault geometries significantly underestimate the area of across-fault juxtaposition relative to realistic fault geometries. In turn this leads to overestimates in the sealing ability of faults, and inaccurate calculation of fault plane properties such as transmissibility multipliers (TMs).

SU‐E‐T‐487: Impact of Geometric Uncertainties in Accelerated Partial Breast Irradiation Using the Strut‐Adjusted Volume Implant (SAVI)

Purpose:Single‐entry multi‐catheter devices have been developed for accelerated partial breast irradiation (APBI). Rotational and translational uncertainties are usually mitigated by quality assurance, however its actual dosimetric impact has not been addressed, which is presented here for SAVI applicator.Methods:Under Institutional Review Board exemption status, we retrospectively analyzed 48 APBI treatment plans using SAVI applicator. For quick calculation of dose‐volume histogram (DVH) considering geometric uncertainties, the coordinate of each voxel of critical organs after rotation or translation along the central catheter was calculated using an in‐house software instead of rotating or translating the entire dose distribution.Results:For most cases, the skin doses increased with rotation to both directions. At 10 degree of rotation, the increase of Dmax (percent prescribed dose) at 50% (median), 75%, 90%, and 100% (maximum) percentile were 0.3%, 1.7%, 5.6% and 25.0%, respectively. The increase of chest wall Dmax at 50%, 75%, 90%, and 100% percentile were 0.1%, 1.0%, 10.3% and 38.8%, respectively. The increase of skin Dmax of the patients with the distance between skin surface and lumpectomy cavity surface ≤16 mm was significantly larger than those with the distance >16 mm. Similarly, patients with the distance between lung and lumpectomy cavity surface ≤20 mm showed higher sensitivity of chest wall Dmax against rotation than patients with the distance <20 mm. Translation of the applicator showed larger impact on the skin dose than rotation, although the effects on chest wall was less than 2% even at 10 mm displacement. At 5 mm of translation, increase of skin Dmax at 50%, 75%, 90%, and 100% percentile were 2.3%, 8.0%, 15.3% and 77.7%, respectively.Conclusion:The impacts of the geometric uncertainties (rotation and translation) of SAVI applicator were investigated. The skin‐lumpectomy cavity and lung‐lumpectomy cavity distances showed significant relationship with skin and chest wall doses, respectively.

Poster - Thurs Eve-13: Modeling the effect of organ motion on cumulative rectal dose using EUD.

A rigid body model and electronic serial portal imaging (EPI) data were used to generate the cumulative dose distribution for the rectum incorporating organ motion during IMRT. The impact of rectal motion was assessed via NTCP and TCP based on equivalent uniform dose per fraction (EUDf ). The rectal positional variations were measured fraction-to-fraction from MV EPI for 20 prostate patients implanted with gold seeds. Five-field (5F) and seven-field (7F) IMRT plans for prostate patients were constructed with prescribed dose 78 Gy/39 fractions using a Pinnacle3 treatment planning system. EUD increased in 45% of the patients with greater than 2.5% increase for 5% of the patients. While EUD decreased in 55% of the patients with greater than 2.5% decrease for 10% of the patients. The amplitudes of EUDf increase and decrease are correlated with the dose gradient. Higher dose gradients lead to higher rectal EUDf change. The rectal NTCP decreases for half of the patients and increases for the other for both 5F and 7F plans. The NTCP decreased with the increasing of dose gradient between the prostate and rectum for 5F and 7F IMRT plans. The correlation coefficient for the rectal NTCP and the dose gradient is - 0.71. The increase or decrease of rectal cumulative dose depends on the dose gradient, motion amplitude and frequency in AP direction. EUDf is a useful QA parameter for interpreting the biological impact of geometric uncertainties on the static dose distribution. Rectal NTCP is patient dependent and must be determined individually.

SU-GG-T-419: IMRT Prostate: Cumulative Rectal Dose Incorporating Rectal Motion

Purpose: For prostate IMRT, a rigid body model and serial portalimaging data were used to generate the cumulative dose distribution for the rectum incorporating internal organ motion by using equivalent uniform dose per fraction (EUDf). The impact of the rectal movement on NTCP over the course of treatment was evaluated. Method and Materials: The rectal positional variations were measured fraction‐to‐fraction from MV portalimaging for 20 prostate patients implanted with gold seeds. The rectal dose change per fraction can be expressed as EUDf variation. Five‐field (5F) and seven‐field (7F) IMRT plans for prostate patients were made with prescribed dose 78 Gy/39 fractions using a Pinnacle3treatment planning system with 15 MV x‐rays. The evaluation was performed by comparing the fractional dose‐volume histogram, the rectal EUDf and NTCP with and without internal organ motion. Results: The increase or decrease of cumulative rectal EUD over the course of treatment depends on the motion amplitude and frequency in the anterior or the posterior directions. EUD increased in 45% of the patients with greater than 2.5% increase for 5% of the patients. EUD decreased in 55% of the patients with greater than 2.5% decrease for 10% of the patients. The rectal NTCP decreases for half of the patients and increases for the other for both 5F and 7F plans in this group. Comparing 5F and 7F plan for the same patient, NTCP is lower for 7F plan compared to 5F plan with and without considering rectal movement. Conclusion: EUDf is a useful QA parameter for interpreting the biological impact of geometric uncertainties on the static dose distribution. The increase or decrease of the cumulative rectal dose depended on the motion amplitude and frequency in the anterior or the posterior direction and rectal NTCP is patient dependent and must be determined individually.

Impact of geometrical uncertainties on 3D CRT and IMRT dose distributions for lung cancer treatment

To quantify the effect of set-up errors and respiratory motion on dose distributions for non-small cell lung cancer (NSCLC) treatment. Irradiations of 5 NSCLC patients were planned with 3 techniques, two (conformal radiation therapy (CRT) and intensity modulated radiation therapy (IMRT1)) with a homogeneous dose in the planning target volume (PTV) and a third (IMRT2) with dose heterogeneity. Set-up errors were simulated for gross target volume (GTV) and organs at risk (OARs). For the GTV, the respiration was also simulated with a periodical motion around a varying average. Two configurations were studied for the breathing motion, to describe the situations of free-breathing (FB) and respiration-correlated (RC) CT scans, each with 2 amplitudes (5 and 10 mm), thus resulting in 4 scenarios (FB_5, FB_10, RC_5 and RC_10). Five thousand treatment courses were simulated, producing probability distributions for the dosimetric parameters. For CRT and IMRT1, RC_5, RC_10 and FB_5 were associated with a small degradation of the GTV coverage. IMRT2 with FB_10 showed the largest deterioration of the GTV dosimetric indices, reaching 7% for Dmin at the 95% probability level. Removing the systematic error due to the periodic breathing motion was advantageous for a 10 mm respiration amplitude. The estimated probability of radiation pneumonitis and acute complication for the esophagus showed limited sensitivity to geometrical uncertainties. Dmax in the spinal cord and the parameters predicting the risk of late esophageal toxicity were associated to a probability up to 50% of violating the dose tolerances. Simulating the effect of geometrical uncertainties on the individual patient plan should become part of the standard pre-treatment verification procedure.

Impact of geometric uncertainties on dose distribution during intensity modulated radiotherapy of head-and-neck cancer: the need for a planning target volume and a planning organ-at-risk volume.

We assessed the effect of geometric uncertainties on target coverage and on dose to the organs at risk (OARS) during intensity-modulated radiotherapy (IMRT) for head-and-neck cancer, and we estimated the required margins for the planning target volume (PTV) and the planning organ-at-risk volume (PRV). For eight head-and-neck cancer patients, we generated IMRT plans with localization uncertainty margins of 0 mm, 2.5 mm, and 5.0 mm. The beam intensities were then applied on repeat computed tomography (CT) scans obtained weekly during treatment, and dose distributions were recalculated.The dose-volume histogram analysis for the repeat ct scans showed that target coverage was adequate (V(100) >/= 95%) for only 12.5% of the gross tumour volumes, 54.3% of the upper-neck clinical target volumes (CTVS), and 27.4% of the lower-neck CTVS when no margins were added for PTV. The use of 2.5-mm and 5.0-mm margins significantly improved target coverage, but the mean dose to the contralateral parotid increased from 25.9 Gy to 29.2 Gy. Maximum dose to the spinal cord was above limit in 57.7%, 34.6%, and 15.4% of cases when 0-mm, 2.5-mm, and 5.0-mm margins (respectively) were used for prv.Significant deviations from the prescribed dose can occur during IMRT treatment delivery for head-and-neck cancer. The use of 2.5-mm to 5.0-mm margins for PTV and PRV greatly reduces the risk of underdosing targets and of overdosing the spinal cord.

Impact of Geometric Uncertainties on Dose Distribution During Intensity Modulated Radiotherapy of Head-and-neck Cancer: The Need for a Planning Target Volume and A Planning Organ-at-Risk Volume

We assessed the effect of geometric uncertainties on target coverage and on dose to the organs at risk (OARS) during intensity-modulated radiotherapy (IMRT) for head-and-neck cancer, and we estimated the required margins for the planning target volume (PTV) and the planning organ-at-risk volume (PRV). For eight headand- neck cancer patients, we generated IMRT plans with localization uncertainty margins of 0 mm, 2.5 mm, and 5.0 mm. The beam intensities were then applied on repeat computed tomography (CT) scans obtained weekly during treatment, and dose distributions were recalculated. The dose–volume histogram analysis for the repeat CT scans showed that target coverage was adequate (V100 ≥ 95%) for only 12.5% of the gross tumour volumes, 54.3% of the upper-neck clinical target volumes (CTVS), and 27.4% of the lower-neck CTVS when no margins were added for PTV. The use of 2.5-mm and 5.0-mm margins significantly improved target coverage, but the mean dose to the contralateral parotid increased from 25.9 Gy to 29.2 Gy. Maximum dose to the spinal cord was above limit in 57.7%, 34.6%, and 15.4% of cases when 0-mm, 2.5-mm, and 5.0-mm margins (respectively) were used for PRV. Significant deviations from the prescribed dose can occur during IMRT treatment delivery for headand- neck cancer. The use of 2.5-mm to 5.0-mm margins for PTV and PRV greatly reduces the risk of underdosing targets and of overdosing the spinal cord.

Evaluation of image-guided radiation therapy (IGRT) technologies and their impact on the outcomes of hypofractionated prostate cancer treatments: A radiobiologic analysis

To quantify the mitigation of geometric uncertainties achieved with the application of various patient setup techniques during the delivery of hypofractionated prostate cancer treatments, using tumor control probability (TCP) and normal tissue complication probability. Five prostate cancer patients with approximately 16 treatment CT studies, taken during the course of their radiation therapy (77 total), were analyzed. All patients were planned twice with an 18 MV six-field conformal technique, with 10- and 5-mm margin sizes, with various hypofractionation schedules (5 to 35 fractions). Subsequently, four clinically relevant patient setup techniques (laser guided and image guided) were simulated to deliver such schedules. As hypothesized, the impact of geometric uncertainties on clinical outcomes increased with more hypofractionated schedules. However, the absolute gain in TCP due to hypofractionation (up to 21.8% increase) was significantly higher compared with the losses due to geometric uncertainties (up to 8.6% decrease). The results of this study suggest that, although the impact of geometric uncertainties on the treatment outcomes increases as the number of fractions decrease, the reduction in TCP due to the uncertainties does not significantly offset the expected theoretical gain in TCP by hypofractionation.

236 Impact of geometrical uncertainties on 3D-CRT and IMRT dose distributions for lung cancer treatment

236 Impact of geometrical uncertainties on 3D-CRT and IMRT dose distributions for lung cancer treatment

Impact of geometric uncertainties on evaluation of treatment techniques for prostate cancer

To assess the impact of patient repositioning and internal organ motion on prostate treatment plans using three-dimensional conformal and intensity-modulated radiotherapy. Four-field, six-field, and simplified intensity-modulated arc therapy plans were generated for 5 prostate cancer patients. The planning target volume was created by adding a 1-cm margin to the clinical target volume. A convolution model was used to estimate the effect of random geometric uncertainties during treatment. Dose statistics, tumor control probabilities, and normal tissue complication probabilities were compared with and without the presence of uncertainty. The impact of systematic uncertainties was also investigated. Compared with the planned treatments, the delivered dose distribution with random geometric uncertainties displayed an increase in the apparent minimal dose to the prostate and seminal vesicles and a decrease in the rectal volume receiving a high dose. This increased the tumor control probabilities and decreased the normal tissue complication probabilities. Changes were seen in the percentage of prostate volume receiving 100% and 95% of the prescribed dose, and the minimal dose and tumor control probabilities for the target volume. In addition, the volume receiving at least 65 Gy, the minimal dose, and normal tissue complication probabilities changed considerably for the rectum. The simplified intensity-modulated arc therapy technique was the most sensitive to systematic errors, especially in the anterior-posterior and superior-inferior directions. Geometric uncertainties should be considered when evaluating treatment plans. Contrary to the widely held belief, increased conformation of the dose distribution is not always associated with increased sensitivity to random geometric uncertainties if a sufficient planning target volume margin is used. Systematic errors may have a variable effect, depending on the treatment technique used.

Modeling geometric uncertainties in radiation therapy

Patient repositioning and organ motion lead to uncertainty in targeting the tumor in radiation therapy. This decreases the dose to the tumor and increases the dose to healthy tissues. Ultimately, tumor control is reduced and complications are increased. This work investigates the hypothesis that modeling geometric uncertainties can accurately estimate the dose distribution delivered when uncertainties are present. These modeling results provide a more accurate representation of the delivered dose distribution than present approaches. These uncertainties are conventionally addressed by adding a margin to the clinical target volume to define a planning target volume (PTV). Despite widespread use, some PTV implementation details have not been addressed. A mathematical model is developed to investigate these details and leads to recommendations for clinical implementation. However, limitations remain when using a PTV. A superior method of accounting for geometric uncertainties incorporates their effect into the dose calculation. This can be achieved by convolution of the planned dose distribution with a probability density function describing the uncertainty. Two assumptions in this model are that the dose distribution is shift invariant and that treatment extends over an infinite number of fractions. The errors resulting from each assumption are quantified. Assuming shift invariance leads to large errors near the patient surface. A “Corrected Convolution” method that reduces these errors was developed. Errors due to finite fractionation are large for very few fractions (hypofractionation), but are unlikely to impact treatment plan evaluation. The impact of geometric uncertainties for hypofractionated prostate cancer treatment is further explored. Hypofractionated treatments were simulated to quantify the impact of geometric uncertainties. The results suggest geometric uncertainties will not limit the clinical effectiveness of prostate hypofractionation. Convolution can assess the sensitivity of different techniques to geometric uncertainties. Simplified intensity modulated arc therapy plans were compared to conventional techniques. The magnitude of the change caused by geometric uncertainties varied by technique. Contrary to common assumptions, geometric uncertainties do not always result in a worse treatment than planned. In conclusion, existing methods to account for geometric uncertainties are limited. Modeling geometric uncertainties with convolution has the potential to improve clinical treatment decisions .