Set Of Contours Research Articles

Multi-camera calibration for accurate geometric measurements in industrial environments

Accurate camera calibration is a challenging task required for 3D reconstruction sensors. In structured light sensors based on laser line projection, it is necessary to determine the position and orientation of the camera relative to the laser plane. The standard approach based on laser plane fitting requires a difficult setup, particularly in multi-camera configurations. This work proposes a calibration method based on a calibration plate with protruding cylinders. The projection of the laser line on the calibration plate produces a set of partial contours of ellipses, resulting from the projection of the cross section of the cylinders. These contours are used to calculate the position and orientation of the camera accurately. The proposed procedure is divided in two steps: a coarse calibration based on a robust estimator, which provides a rough approximation, and a fine calibration that iteratively optimizes the solution. The result is an accurate and robust procedure that can be applied to multiple cameras simultaneously on devices with limited computational power. Extended tests are used to validate the procedure with test pieces. Results indicate the calibration can be performed in 30 ms, producing a calibration error of only 0.027 mm.

Data Integrity Systems for Organ Contours in Radiation Therapy Planning

The purpose of this research is to develop effective data integrity models for contoured anatomy in a radiotherapy workflow for both real-time and retrospective analysis. Within this study, two classes of contour integrity models were developed: data driven models and contiguousness models. The data driven models aim to highlight contours which deviate from a gross set of contours from similar disease sites and encompass the following regions of interest (ROI): bladder, femoral heads, spinal cord, and rectum. The contiguousness models, which individually analyze the geometry of contours to detect possible errors, are applied across many different ROI's and are divided into two metrics: Extent and Region Growing over volume. After analysis, we found that 70% of detected bladder contours were verified as suspicious. The spinal cord and rectum models verified that 73% and 80% of contours were suspicious respectively. The contiguousness models were the most accurate models and the Region Growing model was the most accurate submodel. 100% of the detected noncontiguous contours were verified as suspicious, but in the cases of spinal cord, femoral heads, bladder, and rectum, the Region Growing model detected additional two to five suspicious contours that the Extent model failed to detect. When conducting a blind review to detect false negatives, it was found that all the data driven models failed to detect all suspicious contours. The Region Growing contiguousness model produced zero false negatives in all regions of interest other than prostate. With regards to runtime, the contiguousness via extent model took an average of 0.2 s per contour. On the other hand, the region growing method had a longer runtime which was dependent on the number of voxels in the contour. Both contiguousness models have potential for real-time use in clinical radiotherapy while the data driven models are better suited for retrospective use.

Evaluation of Three Deformable Image Registration Techniques Between CT and CBCT in Prostate Cancer Radiation Therapy

Daily CBCT images can be used to track dose to critical structures in order to identify deviations from the planned treatment, however, this requires accurate contour delineation on daily images and accumulation of the daily dose. Deformable image registration (DIR) between the planning CT (pCT) and CBCT can be used to generate daily contours on the CBCT and to transfer the daily dose for dose accumulation. In this study two novel methods for DIR are compared to a previously described normalized intensity-based DIR. CT and corresponding CBCT images (totaling 124 dose fractions) from 16 subjects treated with radiation therapy for prostate cancer were analyzed. Contours for the bladder, rectum, and prostate were manually defined on each CBCT by a radiation oncologist. For each patient, all contours were deformed from the pCT to each daily CBCT using the following three DIR methods: normalized intensity based (NID), contour surface based (CSD), and multi-modality (MD). The CSD method minimizes surface differences between corresponding contours on two images. The MD method utilizes a diffusion regularized, free-form deformation algorithm to maximize the correspondence of high-dimensional feature descriptors computed for each image voxel. The deformed contours (dPC) were compared to manual reference contours on each input image by calculating dice similarity coefficients (DSC). Paired t-test p-values were used to evaluate statistical significance of the results. Average DSC across all scans and patients for all structures improved from 0.66 for NID to 0.87 for CSD and 0.74 for MD, with an overall improvement of 60.4% and 23.1%, respectively. The CSD dPC were significantly more accurate for prostate, bladder, and rectum over the other two methods (p<1.0*10-15). Though CSD performed best overall, the MD method outperformed NID, improving all three structures by an average of 23%.Abstract TU_39_3279; Table 1Average DSC (Range)NID DICECSD DICEMD DICEBladder0.70 ± 0.14 (0.17-0.93)0.88 ± 0.09 (0.47-0.97)0.78 ± 0.15 (0.28-0.94)Prostate0.67 ± 0.13 (0.28-0.90)0.87 ± 0.05 (0.71-0.96)0.72 ± 0.11 (0.35-0.93)Rectum0.61 ± 0.11 (0.19-0.80)0.84 ± 0.05 (0.72-0.94)0.71 ± 0.08 (0.45-0.88) Open table in a new tab Out of the three methods tested, the CSD method was the most accurate, with average dice falling between 0.84 and 0.88 for all three analyzed structures. However, this method requires contours to be present on both images, meaning it cannot be used for the purpose of adaptive recontouring. It can, however, be used as an adjunct to other methods. The MD method, which does not require two sets of contours, also performed well, with average dice falling between 0.71 and 0.78 for all structures. These two methods could be used in combination to achieve successful and accurate adaptive recontouring and dose transfer for accumulation and dose tracking. We plan to continue to investigate the potential of these novel DIR methods for the purpose of dose-guided adaptive therapy.

Dosimetry Differences between Treatment Plans Computed with Finite Size Pencil Beam Algorithm and Monte Carlo Algorithm Using a Multileaf Collimator Equipped Robotic Radiosurgery System

It has been a constant concern for clinicians on the realistic dose differences computed with an early-available finite size pencil beam algorithm (FSPB) for the treatment plans using a multileaf collimators equipped robotic radiosurgery system (RR-MLC) from the doses computed by an industry well-accepted algorithm, such as Monte Carlo dose algorithm (MC). This study was designed to quantify these dose disparities in the RR-MLC treatment plans for selected anatomic sites in using both FSPB and MC. A total of forty RR-MLC stereotactic ablative radiation therapy (SART) plans computed with FSPB were retrospectively reviewed and compared with MC computed results, including plans for detached lung cancer (or tumors fully surrounded by lung tissues,n=10), non-detached lung cancer (or tumor touched the chest wall or mediastinum, n=10), intracranial(n=10), and pancreas lesions(n=5). MC was commissioned with an uncertainty of 0.2%. All the selected plans were recomputed in the Precision™ planning system for both algorithms with the identical set of contours, optimizations, and MU’s. The dosimetry for planning tumor volume (PTV) and major organ at risks (OAR) was compared with each paired plan. A “normalized deviation” (N_dev=(MC-FSPB)/FSPB) and paired two sample t-tests were used for quantitative analysis. AS shown in Table 1, FSPB plans of RR-MLC overestimates D95 of PTV by averaging 18.9% when compared with the MC computed plans in detached lung cases, and 17.0% in non-detached lung cases. These dose differences appear much less in intracranial and pancreas cases, the differences of D95 of PTV are 1.3% and 2.0% respectively. The lung SART plans encompass the situation with the largest tissue density heterogeneity among all the groups, especially those for detached lung tumors where the disadvantage of FSPB shows largest in dose computations. There is less significant difference (1.3-4.9%) between FSPB and MC calculated maximum dose to OAR’s in each case. The results from this investigation suggest a significant dose difference between lung SART plans computed with FSPB and MC, in which FSPB overestimates D95 of the lung PTV by about 17% to 18.9% and slightly underestimates the hot spots in the surrounding OAR by about 5% or less. The dose differences between plans using the two algorithms are about 2% of less in SART plans for intracranial and pancreatic lesions. These results shall provide a useful guide in evaluating RR-MLC plans when MC is not used.Abstract TU_31_3199; Table 1Dosimetry Comparisons of RR-MLC Plans Computed with FSPB and MC for Selected SitesPTV DosesNon-Detached lungDetached lungPancreasIntracranialN_devp-valueN_devp-valueN_devP-valueN_devp-valueD95-17.0%<0.001-18.9%0.013-1.3%0.128-2.0%0.013Dmean-13.8%<0.001-14.7%0.0626-1.2%0.062-1.1%0.0626Dmax-5.2%<0.001-4.5%<0.0015.1%0.0024.3%<0.001 Open table in a new tab

Automatic Global Level Set Approach for Lumbar Vertebrae CT Image Segmentation.

Vertebrae computed tomography (CT) image automatic segmentation is an essential step for Image-guided minimally invasive spine surgery. However, most of state-of-the-art methods still require human intervention due to the inherent limitations of vertebrae CT image, such as topological variation, irregular boundaries (double boundary, weak boundary), and image noise. Therefore, this paper intentionally designed an automatic global level set approach (AGLSA), which is capable of dealing with these issues for lumbar vertebrae CT image segmentation. Unlike the traditional level set methods, we firstly propose an automatically initialized level set function (AILSF) that comprises hybrid morphological filter (HMF) and Gaussian mixture model (GMM) to automatically generate a smooth initial contour which is precisely adjacent to the object boundary. Secondly, a regularized level set formulation is introduced to overcome the weak boundary leaking problem, which utilizes the region correlation of histograms inside and outside the level set contour as a global term. Ultimately, a gradient vector flow (GVF) based edge-stopping function is employed to guarantee a fast convergence rate of the level set evolution and to avoid level set function oversegmentation at the same time. Our proposed approach has been tested on 115 vertebrae CT volumes of various patients. Quantitative comparisons validate that our proposed AGLSA is more accurate in segmenting lumbar vertebrae CT images with irregular boundaries and more robust to various levels of salt-and-pepper noise.

On coupled-channel dynamics in the presence of anomalous thresholds

We explore a general framework to treat coupled-channel systems in the presence of overlapping left- and right-hand cuts as well as anomalous thresholds. Such systems are studied in terms of a generalized potential, where we exploit the known analytic structure of $t$- and $u$-channel forces as the exchange masses approach their physical values. Given an approximate generalized potential the coupled-channel reaction amplitudes are defined in terms of nonlinear systems of integral equations. For large exchange masses, where there are no anomalous thresholds present, conventional $N/D$ methods are applicable to derive numerical solutions to the latter. At a formal level a generalization to the anomalous case is readily formulated by use of suitable contour integrations with amplitudes to be evaluated at complex energies. However, it is a considerable challenge to find numerical solutions to anomalous systems set up on a set of complex contours. By suitable deformations of left-hand and right-hand cut lines we manage to establish a framework of linear integral equations defined for real energies. Explicit expressions are derived for the driving terms that hold for an arbitrary number of channels. Our approach is illustrated in terms of schematic three-channel systems. It is demonstrated that despite the presence of anomalous thresholds the scattering amplitude can be represented in terms of three phase shifts and three inelasticity parameters, as one would expect from the coupled-channel unitarity condition.

Equi-Coverage Contours in Cellular Networks

In this letter, we introduce a general cellular network model, where: 1) users and base stations (BSs) are distributed as two general point processes that may be coupled; 2) pathloss is assumed to follow a multi-slope power-law pathloss model; and 3) fading (power) is assumed to be independent across all wireless links. For this setup, we first obtain a set of contours representing the same meta distribution of signal-to-interference ratio , which is the distribution of the conditional coverage probability given the point process, for different values of the parameters of the pathloss function and BS and user point processes. This general result is then specialized to 3GPP-inspired user and BS configurations obtained by combining Poisson point process and Poisson cluster process.

A novel perimeter-based index for evaluating the similarity of structure contours

AbstractIntroductionThe authors proposed a novel perimeter-based index (PBI) that was capable of evaluating the accuracy in the appraisal of auto-segmentation software. A quantitative value, that is time saved in editing the auto-segmented contours, was used to compare the effectiveness of two other commonly used indices in this study.MethodsThe relationship between the proposed index and the amount of the contouring time that could be saved was studied. The performances of two other commonly used similarity indices, namely Dice similarity coefficient (DSC) and the modified normalised average Hausdorff distance (MNAHD), were also evaluated. Ten nasopharyngeal cases and ten prostate cases that were previously treated with intensity-modulated radiation therapy technique were recruited as the validation cases in this study. Three observers were invited to contour four structures (bladder, rectum, brain stem and parotid gland) on computed tomography images of the validation cases without any aids. The time taken for contouring was recorded as the manual contouring time. By using an atlas-based auto-segmentation software, three sets of contours were generated for each validation case with different library sizes to produce different degrees of similarity level. The values of the three similarity indices of the auto-segmented contours were calculated. The observers were asked to edit the auto-segmented contours and the editing time was recorded.The correlation between the editing time and the similarity indices was studied. The amount of time saved was calculated by subtracting the editing time from the manual contouring time. The performances of PBI, DSC and MNAHD were evaluated using Pearson correlation coefficient and receiver operating curve (ROC) analysis.ResultsThe PBI showed a positive linear relationship with the amount of contouring time saved. Pearson correlation coefficient ranged from 0·73 to 0·86 for the four structures. The PBI had a stronger correlation than the DSC in bladder and parotid gland, while there was no significant difference between the two indices in rectum and brain stem. The MNAHD had an inferior correlation than the proposed index. For the ROC analysis, the cut-off values for the PBI were 0·549, 0·401 and 0·301 for the three levels of contouring time saved, namely 50, 25 and 0%, respectively. The accuracy of PBI was over 77% and the Youden index was &gt;0·6 for all three levels.ConclusionsThe proposed index showed a stronger relationship to the amount of contouring time saved. It was a simple tool that could be used to evaluate the performance of different segmentation algorithms.

Level sets-based image segmentation approach using statistical shape priors

A robust 3-D segmentation technique incorporated with the level sets concept and based on both shape and intensity constraints is introduced. A partial differential equation (PDE) is derived to describe the evolution of the level set contours. This PDE does not contain weighting parameters that need to be tuned, which overcomes the drawbacks of other PDE approaches. The shape information is collected from a set of co-aligned manually segmented contours of the training data. A promising statistical approach is used to get the distribution of the intensity gray values. The introduced statistical approach is built by modeling the empirical PDF (normalized histogram of occurrences) for the intensity level distribution with a linear combination of Gaussians (LCG) incorporating both negative and positive components. An Expectation-Maximization (EM) algorithm is modified to deal with the LCGs, and we also proposed an EM-based sequential technique to acquire a close initial LCG approximation for the modified EM algorithm to start with. The PDF of the intensity levels is incorporated in the speed function of the moving level set to specify the evolution direction. Experimental results show how accurately the approach is in segmenting various types of 2-D and 3-D datasets comprising medical images.

An Iterative Diffeomorphic Algorithm for Registration of Subdivision Surfaces: Application to Congenital Heart Disease.

In this paper, we present a new diffeomorphic registration algorithm for the registration of 3D models to 3D points. A biventricular template is iteratively fitted to the data by a series of implicitly constrained diffeomorphic linear least squares fits with decreasing regularization weights before performing an explicitly constrained diffeomorphic fit. The algorithm has been tested on a set of manual contours from 20 patients with a variety of congenital heart disease. Registration accuracy was assessed by calculating the mean point-to-point distance and the Dice overlap metric. Results showed that the method was able to accurately fit the biventricular model to 3D points and that the deformable model was able to fit all the pathologies while being diffeomorphic. The algorithm took approximately 5 minutes to fit each case, with an average of 52,580 points per case.

Data integrity systems for organ contours in radiation therapy planning.

The purpose of this research is to develop effective data integrity models for contoured anatomy in a radiotherapy workflow for both real‐time and retrospective analysis. Within this study, two classes of contour integrity models were developed: data driven models and contiguousness models. The data driven models aim to highlight contours which deviate from a gross set of contours from similar disease sites and encompass the following regions of interest (ROI): bladder, femoral heads, spinal cord, and rectum. The contiguousness models, which individually analyze the geometry of contours to detect possible errors, are applied across many different ROI's and are divided into two metrics: Extent and Region Growing over volume. After analysis, we found that 70% of detected bladder contours were verified as suspicious. The spinal cord and rectum models verified that 73% and 80% of contours were suspicious respectively. The contiguousness models were the most accurate models and the Region Growing model was the most accurate submodel. 100% of the detected noncontiguous contours were verified as suspicious, but in the cases of spinal cord, femoral heads, bladder, and rectum, the Region Growing model detected additional two to five suspicious contours that the Extent model failed to detect. When conducting a blind review to detect false negatives, it was found that all the data driven models failed to detect all suspicious contours. The Region Growing contiguousness model produced zero false negatives in all regions of interest other than prostate. With regards to runtime, the contiguousness via extent model took an average of 0.2 s per contour. On the other hand, the region growing method had a longer runtime which was dependent on the number of voxels in the contour. Both contiguousness models have potential for real‐time use in clinical radiotherapy while the data driven models are better suited for retrospective use.

The impact of contour variation on tumour control probability in anal cancer

BackgroundWhile intensity modulated radiotherapy (IMRT) has been widely adopted for the treatment of anal cancer (AC), the added contour complexity poses potential risks. This study investigates the impact of contour variation on tumour control probability (TCP) when using IMRT for AC.MethodsNine Australian centres contoured a single computed tomography dataset of a patient with AC. The same optimised template-based IMRT planning protocol was applied to each contour set to generate nine representative treatment plans and their corresponding dose volume histograms. A geometric analysis was performed on all contours. The TCP was calculated for each plan using the linear quadratic and logitEUD model.ResultsThe median concordance index (CI) for the bladder, head and neck of femur, bone marrow, small bowel and external genitalia was 0.94, 0.88, 0.84, 0.65 and 0.65, respectively. The median CI for the involved nodal, primary tumour and elective clinical target volumes were 0.85, 0.77 and 0.71, respectively. Across the nine plans, the TCP was not significantly different. Variation in TCP between plans increased as tumour cell load increased or radiation dose decreased.ConclusionsWhen using IMRT for AC, contour variations generated from a common protocol within the limits of minor deviations do not appear to have a significant impact on TCP. Contouring variations may be more critical with increasing tumour cell load or reducing radiotherapy dose.

Automatic segmentation of chondroblastoma from X-ray images using active contour and levelset method

Chondroblastoma is a benign but locally aggressive bone tumor found usually in the age below 25 years. Chondroblastoma is a destructive type of lesion with a thin radio dense border which is normally seen in the epiphysis of long bones. The benign tumors have similarities in pathology and could be related with histogenic similarity. This tumor reduces the strength of affected bone and may leads to death if not treated early. Chondroblastoma can be diagnosed from X-ray/CT/MRI images and the treatment is its removal by surgical methods. Diagnosis of Chondroblastoma is difficult due to the similarities with other benign tumors like chondromyxoid fibroma. To reduce diagnostic errors, computer aided methods can adopt. This work focuses on automatic segmentation of Chondroblastoma using active contour and level set method which gives better segmentation results and a mild stone to CAD design.

Deformable image registration–based contour propagation yields clinically acceptable plans for MRI-based cervical cancer brachytherapy planning

PurposeTo study the dosimetric impact of deformable image registration–based contour propagation on MRI-based cervical cancer brachytherapy planning. Methods and MaterialsHigh-risk clinical target volume (HRCTV) and organ-at-risk (OAR) contours were delineated on MR images of 10 patients who underwent ring and tandem brachytherapy. A second set of contours were propagated using a commercially available deformable registration algorithm. “Manual-contour” and “propagated-contour” plans were optimized to achieve a maximum dose to the most minimally exposed 90% of the volume (D90) (%) of 6 Gy/fraction, respecting minimum dose to the most exposed 2cc of the volume (D2cc) OAR constraints of 5.25 Gy and 4.2 Gy/fraction for bladder and rectum/sigmoid (86.5 and 73.4 Gy equivalent dose in 2 Gy fractions [EQD2] for external beam radiotherapy [EBRT] + brachytherapy, respectively). Plans were compared using geometric and dosimetric (total dose [EQD2] EBRT + brachytherapy) parameters. ResultsThe differences between the manual- and propagated-contour plans with respect to the HRCTV D90 and bladder, rectum, and sigmoid D2cc were not statistically significant (per-fraction basis). For the EBRT + brachytherapy course, the D2cc delivered to the manually contoured OARs by the propagated-contour plans ranging 98–107%, 95–105%, and 92–108% of the dose delivered by the manual-contour plans (max 90.4, 70.3, and 75.4 Gy for the bladder, rectum, and sigmoid, respectively). The HRCTV dose in the propagated-contour plans was 97–103% of the dose in the manual-contour plans (maximum difference 2.92 Gy). Increased bladder filling resulted in increased bladder dose in manual- and propagated-contour plans. ConclusionsWhen deformable image registration–propagated contours are used for cervical brachytherapy planning, the HRCTV dose is similar to the dose delivered by manual-contour plans and the doses delivered to the OARs are clinically acceptable, suggesting that our algorithm can replace manual contouring for appropriately selected cases that lack major interfractional anatomical changes.

Dosimetric Effects of Head and Neck Immobilization Devices on Multi-field Intensity Modulated Radiation Therapy for Nasopharyngeal Carcinoma.

Background: In practice, the dose perturbation effect of head and neck immobilization devices is often overlooked in intensity-modulated radiation therapy (IMRT) for nasopharyngeal carcinoma (NPC). Purpose of this study is to verify and analyze the dosimetric effect of head and neck immobilization devices on NPC multi-field IMRT.Methods: Ten patients with nasopharyngeal carcinoma were randomly selected. Two sets of body contours were established for each patient. One set of body contours did not contain the immobilization device, and the other contour set included the immobilization device. For each patient, dose calculations were conducted for the two sets of contours using the same 9-field IMRT plan, which were recorded as Plan- and Plan+. The dose difference caused by the head and neck immobilization devices was assessed by comparing the dose-volume histogram (DVH) parameter results and by plan subtraction. The gafchromic EBT3 film and anthropomorphic phantom were used to verify the calculated doses.Results: The target coverage and average dose of Plan+ were lower than those of Plan- : the prescription dose coverage rates for PTVnx, PTVnd, PTV1 and PTV2 decreased by 2.4%, 9.9%, 1.5%, and 3.6%, respectively, and the mean doses were reduced by 0.9%, 1.9%, 1.1%, and 1.5%, respectively. Doses in the organs at risk showed no significant differences or slight reductions (the maximum reduction in mean dose was 1.7%). From the EBT3 measurements, the skin dose on the posterior neck was increased by approximately 53%.Conclusion: The attenuation and bolus effects of the head and neck immobilization device reduce dose coverage rate and average dose of the planning target volumes in nasopharyngeal carcinoma and lead to an increase in the skin dose. During treatment planning and dose calculation, the immobilization device should be included within body contour to account for the dose attenuation and skin dose increment.

Explosive wave propagation in the presence of antiseismic protective curtain

The objective of this work is to study theoretically the stress wave formation in the borehole-to-borehole space and on the boundary with the borehole contour set by creating a directed fracture system in it. The features of an interaction of the weak stress wave at a large-scale blasting in the short-delay blast mode (SDB) with the massif and the protective contour set of boreholes, are considered. It is noted the effect of the contour set parameters on wave processes. The constructive solutions are proposed in order to reduce the dynamic loadings in the peripheral rock massif. The proposals consist in applying a combined construction of garland charges to form the seismo protective structure that consists of a contour fracture and one-sided disrupted zone within the boundaries of the blasted block.

Dosimetric comparison of sector-blocked and non–sector-blocked Gamma Knife Perfexion treatment plans for trigeminal neuralgia

The purpose of this study is to compare the quality of trigeminal neuralgia (TN) treatment plans with and without utilizing sector blocking. Twelve patients with 13 cases of TN were evaluated in this retrospective study. Identical magnetic resonance imaging (MRI) contour sets and prescription doses used in treatments were reused for all plans. Treatment plans were compared on the basis of the amount of dose received by critical structures (i.e., brainstem, ipsilateral temporal lobe, optic chiasm, optic nerves, cochlea) and the estimated total treatment time. The use of sector blocking resulted in a statistically significant decrease in the radiation dose to the brainstem but increased doses to the cerebellum and temporal lobe regions. The magnitude of these differences was small and individual patient anatomy specific. The use of sector blocking also resulted in a statistically significant increase in the treatment time. The magnitude of the change in treatment time was dependent on the number of sectors blocked. Our study suggests some potential benefits, as well as disadvantages, to the use of sector blocking in the treatment of TN. Treatment decision needs to be individualized based on the patient's anatomy and overall condition.

Cross-cultural empirical aesthetics.

Cross-cultural empirical aesthetics seeks to determine whether the psychological processes underlying aesthetic preference are universal. Here we provide a critical review of the field's origin, development, and current state. Our goal is to evaluate the evidence and separate what is actually known from what is only assumed. We conclude that the evidence shows that people from different cultures base their aesthetic preference on a common set of formal features, including symmetry, complexity, proportion, contour, brightness, and contrast. The reason for this commonality is that aesthetic preference emerges from basic perceptual and valuation processes that are common to all humans, and to many other animals.

Interobserver reliability of computed tomographic contouring of canine tonsils in radiation therapy treatment planning.

In radiation therapy (RT) treatment planning for canine head and neck cancer, the tonsils may be included as part of the treated volume. Delineation of tonsils on computed tomography (CT) scans is difficult. Error or uncertainty in the volume and location of contoured structures may result in treatment failure. The purpose of this prospective, observer agreement study was to assess the interobserver agreement of tonsillar contouring by two groups of trained observers. Thirty dogs undergoing pre- and post-contrast CT studies of the head were included. After the pre- and postcontrast CT scans, the tonsils were identified via direct visualization, barium paste was applied bilaterally to the visible tonsils, and a third CT scan was acquired. Data from each of the three CT scans were registered in an RT treatment planning system. Two groups of observers (one veterinary radiologist and one veterinary radiation oncologist in each group) contoured bilateral tonsils by consensus, obtaining three sets of contours. Tonsil volume and location data were obtained from both groups. The contour volumes and locations were compared between groups using mixed (fixed and random effect) linear models. There was no significant difference between each group's contours in terms of three-dimensional coordinates. However there was a significant difference between each group's contours in terms of the tonsillar volume (P<0.0001). Pre- and postcontrast CT can be used to identify the location of canine tonsils with reasonable agreement between trained observers. Discrepancy in tonsillar volume between groups of trained observers may affect RT treatment outcome.

Pancreatic gross tumor volume contouring on computed tomography (CT) compared with magnetic resonance imaging (MRI): Results of an international contouring conference

PurposeAccurate identification of the gross tumor volume (GTV) in pancreatic adenocarcinoma is challenging. We sought to understand differences in GTV delineation using pancreatic computed tomography (CT) compared with magnetic resonance imaging (MRI). Methods and materialsTwelve attending radiation oncologists were convened for an international contouring symposium. All participants had a clinical and research interest in pancreatic adenocarcinoma. CT and MRI scans from 3 pancreatic cases were used for contouring. CT and MRI GTVs were analyzed and compared. Interobserver variability was compared using Dice's similarity coefficient (DSC), Hausdorff distances, and Jaccard indices. Mann-Whitney tests were used to check for significant differences. Consensus contours on CT and MRI scans and constructed count maps were used to visualize the agreement. Agreement regarding the optimal method to determine GTV definition using MRI was reached. ResultsSix contour sets (3 from CT and 3 from MRI) were obtained and compared for each observer, totaling 72 contour sets. The mean volume of contours on CT was significantly larger at 57.48 mL compared with a mean of 45.76 mL on MRI, P = .011. The standard deviation obtained from the CT contours was significantly larger than the standard deviation from the MRI contours (P = .027). The mean DSC was 0.73 for the CT and 0.72 for the MRI (P = .889). The conformity index measurement was similar for CT and MRI (P = .58). Count maps were created to highlight differences in the contours from CT and MRI. ConclusionsUsing MRI as a primary image set to define a pancreatic adenocarcinoma GTV resulted in smaller contours compared with CT. No differences in DSC or the conformity index were seen between MRI and CT. A stepwise method is recommended as an approach to contour a pancreatic GTV using MRI.