Computed Tomography Contours Research Articles

Training deep-learning segmentation models from severely limited data.

To enable generation of high-quality deep learning segmentation models from severely limited contoured cases (e.g., ~10 cases). Thirty head and neck computed tomography (CT) scans with well-defined contours were deformably registered to 200 CT scans of the same anatomic site without contours. Acquired deformation vector fields were used to train a principal component analysis (PCA) model for each of the 30 contoured CT scans by capturing the mean deformation and most prominent variations. Each PCA model can produce an infinite number of synthetic CT scans and corresponding contours by applying random deformations. We used 300, 600, 1000, and 2000 synthetic CT scans and contours generated from one PCA model to train V-Net, a 3D convolutional neural network architecture, to segment parotid and submandibular glands. We repeated the training using same numbers of training cases generated from 7, 10, 20, and 30 PCA models, with the data distributed evenly between each PCA model. Performance of the segmentation models was evaluated with Dice similarity coefficients between auto-generated contours and physician-drawn contours on 162 test CT scans for parotid glands and another 21 test CT scans for submandibular glands. Dice values varied with the number of synthetic CT scans and the number of PCA models used to train the network. By using 2000 synthetic CT scans generated from 10 PCA models, we achieved Dice values of 82.8%±6.8% for right parotid, 82.0%±6.9% for left parotid, and 74.2%±6.8% for submandibular glands. These results are comparable with those obtained from state-of-the-art auto-contouring approaches, including a deep learning network trained from more than 1000 contoured patients and a multi-atlas algorithm from 12 well-contoured atlases. Improvement was marginal when >10 PCA models or >2000 synthetic CT scans were used. We demonstrated an effective data augmentation approach to train high-quality deep learning segmentation models from a limited number of well-contoured patient cases.

Hippocampus segmentation in CT using deep learning: impact of MR versus CT-based training contours.

.Purpose: Hippocampus contouring for radiotherapy planning is performed on MR image data due to poor anatomical visibility on computed tomography (CT) data. Deep learning methods for direct CT hippocampus auto-segmentation exist, but use MR-based training contours. We investigate if these can be replaced by CT-based contours without loss in segmentation performance. This would remove the MR not only from inference but also from training.Approach: The hippocampus was contoured by medical experts on MR and CT data of 45 patients. Convolutional neural networks (CNNs) for hippocampus segmentation on CT were trained on CT-based or propagated MR-based contours. In both cases, their predictions were evaluated against the MR-based contours considered as the ground truth. Performance was measured using several metrics, including Dice score, surface distances, and contour Dice score. Bayesian dropout was used to estimate model uncertainty.Results: CNNs trained on propagated MR contours (median Dice 0.67) significantly outperform those trained on CT contours (0.59) and also experts contouring manually on CT (0.59). Differences between the latter two are not significant. Training on MR contours results in lower model uncertainty than training on CT contours. All contouring methods (manual or CNN) on CT perform significantly worse than a CNN segmenting the hippocampus directly on MR (median Dice 0.76). Additional data augmentation by rigid transformations improves the quantitative results but the difference remains significant.Conclusions: CT-based training contours for CT hippocampus segmentation cannot replace propagated MR-based contours without significant loss in performance. However, if MR-based contours are used, the resulting segmentations outperform experts in contouring the hippocampus on CT.

Technical Note: Efficient and accurate MRI-only based treatment planning of the prostate using bulk density assignment through atlas-based segmentation.

This study investigates the dosimetric accuracy as well as the robustness of a bulk density assignment approach to magnetic resonance imaging (MRI)-only based treatment planning of the prostate, with bulk density regions automatically identified using atlas-based segmentation (ABS). Twenty prostate radiotherapy patients received planning computed tomography (CT) and MRI scans and were treated with volumetric modulated arc therapy (VMAT). Two bulk densities were set, one for bone and one for soft tissue. The bone contours were created by using ABS followed by manual modification if considered necessary. A range of soft tissue and bone density pairs, between 0.95 and 1.03g/cm3 with increments of 0.01 for soft tissue, and between 1.15 and 1.65g/cm3 with increments of 0.05 for bone, were evaluated. Using the density pair giving the lowest dose difference compared to the CT-based dose, dose differences were calculated using both the manually modified bone contours and the bone contours from ABS. Contour overlap measurements between the ABS contours and the manually modified contours were calculated. The dose comparison shows a very good agreement with the CT when using 0.98g/cm3 for soft tissue and 1.20g/cm3 for bone, with a dose difference less than 1 % in average dose in all regions of interest. The mean Dice similarity coefficient for bone was 0.94 and the Mean Distance to Agreement was <1mm in most cases. Using bulk density assignment on MR images with suitable densities for bone and soft tissue results in clinically acceptable dose differences compared to dose calculated on the CT, for both atlas-based and manual bone contours. This demonstrates that an integrated MRI-only pathway utilizing a bulk density assignment for two tissue types is a feasible and robust approach for patients with prostate cancer treated with VMAT.

Development and clinical implementation of eclipse scripting-based automated patient-specific collision avoidance software.

PurposeIncreased use of Linac‐based stereotactic radiosurgery (SRS), which requires highly noncoplanar gantry trajectories, necessitates the development of efficient and accurate methods of collision detection during the treatment planning process. This work outlines the development and clinical implementation of a patient‐specific computed tomography (CT) contour‐based solution that utilizes Eclipse Scripting to ensure maximum integration with clinical workflow.MethodsThe collision detection application uses triangle mesh structures of the gantry and couch, in addition to the body contour of the patient taken during CT simulation, to virtually simulate patient treatments. Collision detection is performed using Binary Tree Hierarchy detection methods. Algorithm accuracy was first validated for simple cuboidal geometry using a calibration phantom and then extended to an anthropomorphic phantom simulation by comparing the measured minimum distance between structures to the predicted minimum distance for all allowable orientations. The collision space was tested at couch angles every 15° from 90 to 270 with the gantry incremented by 5° through the maximum trajectory. Receiver operating characteristic curve analysis was used to assess algorithm sensitivity and accuracy for predicting collision events. Following extensive validation, the application was implemented clinically for all SRS patients.ResultsThe application was able to predict minimum distances between structures to within 3 cm. A safety margin of 1.5 cm was sufficient to achieve 100% sensitivity for all test cases. Accuracy obtained was 94.2% with the 5 cm clinical safety margin with 100% true positive collision detection. A total of 88 noncoplanar SRS patients have been currently tested using the application with one collision detected and no undetected collisions occurring. The average time for collision testing per patient was 2 min 58 s.ConclusionsA collision detection application utilizing patient CT contours was developed and successfully clinically implemented. This application allows collisions to be detected early during the planning process, avoiding patient delays and unnecessary resource utilization if detected during delivery.

Computed Tomography Myelosimulation Versus Magnetic Resonance Imaging Registration to Delineate the Spinal Cord During Spine Stereotactic Radiosurgery

Underestimation of the spinal cord's volume or position during spine stereotactic radiosurgery can lead to severe myelopathy, whereas overestimation can lead to tumor underdosage. Spinal cord delineation is commonly achieved by registering a magnetic resonance imaging (MRI) study with a computed tomography (CT) simulation scan or by performing myelography during CT simulation (myelosim). We compared treatment planning outcomes for these 2 techniques. Twenty-three cases of spine stereotactic radiosurgery were analyzed that had both a myelosim and corresponding MRI study for registration. The spinal cord was contoured on both imaging data sets by 2 independent blinded physicians, and Dice similarity coefficients were calculated to compare their spatial overlap. Two treatment plans (16 Gy and 18 Gy) were created using the MRI and CT contours (92 plans total). Dosimetric parameters were extracted and compared by modality to assess tumor coverage and spinal cord dose. No differences were found in the partial spinal cord volumes contoured on MRI versus myelosim (4.71 ± 1.09 vs. 4.55 ± 1.03 cm3; P= 0.34) despite imperfect spatial agreement (mean Dice similarity coefficient, 0.68 ± 0.05). When the registered MRI contours were used for treatment planning, significantly worse tumor coverage and greater spinal cord doses were found compared with myelosim planning. For the 18-Gy plans, 10 of 23 MRI cases (43%) exceeded the spinal cord or cauda dose constraints when using myelosim as the reference standard. Significant spatial, rather than volumetric, differences were found between the MRI- and myelosim-defined spinal cord structures. Tumor coverage was compromised with MRI-based planning, and the high spinal cord doses were a concern. Future work is necessary to compare thin-cut, volumetric MRI registration or MRI simulation with myelosim.

A Prospective Comparison of Computed Tomography with Transrectal Ultrasonography Assistance and Magnetic Resonance Imaging–Based Target-Volume Definition During Image Guided Adaptive Brachytherapy for Cervical Cancers

Although magnetic resonance imaging (MRI) represents the gold standard for image guided adaptive brachytherapy (IGABT) for cervical cancer, the majority of brachytherapy (BT) continues to be guided by computed tomography (CT). However, CT seems to overestimate the target-volume definition, and the potential of transrectal ultrasonography (TRUS) needs further evaluation. This prospective, comparative study aimed to evaluate CT-based target contouring with the incorporation of TRUS during BT. Patients with locally advanced cervical cancer undergoing magnetic resonance IGABT between January 2013 and March 2014 were included. During the BT procedure, TRUS imaging with central tandem in situ was acquired at 3 representative levels. Reference points/dimensions (D1-D4) of the hypoechoic region on TRUS images with respect to the central tandem were recorded. CT and magnetic resonance BT planning imaging was performed after BT application. The high-risk clinical target volume (HR-CTV) was contoured on CT scans with incorporation of clinical and TRUS imaging findings and was compared with the gold standard MRI-based target approach. The image sets of 25 patients (International Federation of Gynecology and Obstetrics clinical stage IIB [11; 44%] and IIIB [14; 56%]) were evaluable. The mean (±standard deviation) volumes of HR-CTV on CT and MRI imaging were 39.1 (±20) cm³ and 39 (±19) cm³, respectively (r = 0.92; P < .001). A significant correlation was found between the HR-CTV dimensions (width and thickness) of CT and MRI scans at various levels (r = 0.70-0.80; P < .001). In addition, the absolute differences in target dimensions between CT and MRI were <0.5 cm. A strong correlation was seen between CT and MRI for patients with medial and lateral parametrial invasion (P < .05) compared with no parametrial disease at BT. Furthermore, the mean differences in HR-CTV width between CT and MRI contours at various levels, irrespective of parametrial involvement, was only 0.1 to 0.4 cm. This study suggests that a CT-based target and organ-at-risk delineation using MRI at diagnosis and real-time TRUS information during BT seems comparable with the gold standard MRI-based approach in IGABT for cervical cancer.

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.

A rapid, computational approach for assessing interfraction esophageal motion for use in stereotactic body radiation therapy planning

PurposeWe present a rapid computational method for quantifying interfraction motion of the esophagus in patients undergoing stereotactic body radiation therapy on a magnetic resonance (MR) guided radiation therapy system. Methods and materialsPatients who underwent stereotactic body radiation therapy had simulation computed tomography (CT) and on-treatment MR scans performed. The esophagus was contoured on each scan. CT contours were transferred to MR volumes via rigid registration. Digital Imaging and Communications in Medicine files containing contour points were exported to MATLAB. In-plane CT and MR contour points were spline interpolated, yielding boundaries with centroid positions, CCT and CMR. MR contour points lying outside of the CT contour were extracted. For each such point, BMR(j), a segment from CCT intersecting BMR(j), was produced; its intersection with the CT contour, BCT(i), was calculated. The length of the segment Sij, between BCT(i) and BMR(j), was found. The orientation θ was calculated from Sij vector components:θ = arctan[(Sij)y / (Sij)x]A set of segments {Sij} was produced for each slice and binned by quadrant with 0° < θ ≤ 90°, 90° < θ ≤ 180°, 180° < θ ≤ 270°, and 270° < θ ≤ 360° for the left anterior, right anterior, right posterior, and left posterior quadrants, respectively. Slices were binned into upper, middle, and lower esophageal (LE) segments. ResultsSeven patients, each having 3 MR scans, were evaluated, yielding 1629 axial slices and 84,716 measurements. The LE segment exhibited the greatest magnitude of motion. The mean LE measurements in the left anterior, left posterior, right anterior, and right posterior were 5.2 ± 0.07 mm, 6.0 ± 0.09 mm, 4.8 ± 0.08 mm, and 5.1 ± 0.08 mm, respectively. There was considerable interpatient variability. ConclusionsThe LE segment exhibited the greatest magnitude of mobility compared with the middle and upper esophageal segments. A novel computational method enables personalized, nonuniform esophageal margins to be tailored to individual patients.

Three-dimensional finite-element based deformable image registration for evaluation of pleural cavity irradiation during photodynamic therapy.

Photodynamic therapy (PDT) is used after surgical resection to treat the microscopic disease for malignant pleural mesothelioma and to increase survival rates. As accurate light delivery is imperative to PDT efficacy, the deformation of the pleural volume during the surgery is studied on its impact on the delivered light fluence. In this study, a three-dimensional finite element-based (3D FEM) deformable image registration is proposed to directly match the volume of lung to the volume of pleural cavity obtained during PDT to have accurate representation of the light fluence accumulated in the lung, heart and liver (organs-at-risk) during treatment. A wand, comprised of a modified endotrachial tube filled with Intralipid and an optical fiber inside the tube, is used to deliver the treatment light. The position of the treatment is tracked using an optical tracking system with an attachment comprised of nine reflective passive markers that are seen by an infrared camera-based navigation system. This information is used to obtain the surface contours of the plural cavity and the cumulative light fluence on every point of the cavity surface that is being treated. The lung, heart, and liver geometry are also reconstructed from a series of computed tomography (CT) scans of the organs acquired in the same patient before and after the surgery. The contours obtained with the optical tracking system and CTs are imported into COMSOL Multiphysics, where the 3D FEM-based deformable image registration is obtained. The delivered fluence values are assigned to the respective positions (x, y, and z) on the optical tracking contour. The optical tracking contour is considered as the reference, and the CT contours are used as the target, which will be deformed. The data from three patients formed the basis for this study. The physical correspondence between the CT and optical tracking geometries, taken at different times, from different imaging devices was established using the 3D FEM-based image deformable registration. The volume of lung was matched to the volume of pleural cavity and the distribution of light fluence on the surface of the heart, liver and deformed lung volumes was obtained. The method used is appropriate for analyzing problems over complicated domains, such as when the domain changes (as in a solid-state reaction with a moving boundary), when the desired precision varies over the entire domain, or when the solution lacks smoothness. Implementing this method in real-time for clinical applications and in situ monitoring of the under- or over- exposed regions to light during PDT can significantly improve the treatment for mesothelioma.

SU‐F‐I‐50: Finite Element‐Based Deformable Image Registration of Lung and Heart

Purpose:Photodynamic therapy (PDT) is used after surgical resection to treat the microscopic disease for malignant pleural mesothelioma and to increase survival rates. Although accurate light delivery is imperative to PDT efficacy, the deformation of the pleural volume during the surgery impacts the delivered light dose. To facilitate treatment planning, we use a finite‐element‐based (FEM) deformable image registration to quantify the anatomical variation of lung and heart volumes between CT pre‐(or post‐) surgery and surface contours obtained during PDT using an infrared camera‐based navigation system (NDI).Methods:NDI is used during PDT to obtain the information of the cumulative light fluence on every cavity surface point that is being treated. A wand, comprised of a modified endotrachial tube filled with Intralipid and an optical fiber inside the tube, is used to deliver the light during PDT. The position of the treatment is tracked using an attachment with nine reflective passive markers that are seen by the NDI system. Then, the position points are plotted as three‐dimensional volume of the pleural cavity using Matlab and Meshlab. A series of computed tomography (CT) scans of the lungs and heart, in the same patient, are also acquired before and after the surgery. The NDI and CT contours are imported into COMSOL Multiphysics, where the FEM‐based deformable image registration is obtained. The NDI and CT contours acquired during and post‐PDT are considered as the reference, and the Pre‐PDT CT contours are used as the target, which will be deformed.Results:Anatomical variation of the lung and heart volumes, taken at different times from different imaging devices, was determined by using our model. The resulting three‐dimensional deformation map along x, y and z‐axes was obtained.Conclusion:Our model fuses images acquired by different modalities and provides insights into the variation in anatomical structures over time.

TU‐H‐CAMPUS‐JeP2‐01: Inter‐Observer Delineation Comparison of Visible Glandular Breast Tissue On Magnetic Resonance Imaging and Computed Tomography (prone and Supine)

Purpose:Breast cancers predominantly arise from Glandular Breast Tissue (GBT). If the GBT can be treated effectively post‐operatively utilising radiotherapy this may be adequate volumetric coverage for adjuvant breast radiotherapy. Adequate imaging of the GBT is necessary and will be assessed between MRI and CT modalities. GBT visualisation is acknowledged to be qualitatively superior on Magnetic Resonance Image (MRI) compared to Computed Tomography (CT), the current radiotherapy imaging standard, however this has not been quantitatively assessed. For radiotherapy purposes it is important that any treatment volume can be consistently defined between observers. This study investigates the consistency of CT and MRI GBT contours for potential radiotherapy planning.Methods:Ten experts (9 breast radiation oncologists and 1 radiologist) contoured the extent of the visible GBT for 33 patients on MRI and CT (both without contrast), which was performed according to a contouring guideline in supine and prone patient positions. The GBT volume was not a conventional whole breast radiotherapy planning volume, but rather the extent of GBT that was indicated from the CT or MR imaging. Volumes were compared utilizing the dice similarity coefficient (DSC), kappa statistic, and Hausdorff Distances (HDs) to ascertain the modality that was most consistently volumed.Results:The inter‐observer concordance was of substantial agreement (kappa above 0.6) for the CT supine, CT prone, MRI supine and MRI prone datasets. The MRI GBT volumes were larger than the CT GBT volumes (p&lt;0.001). Inter‐observer conformity was higher for CT than MRI, although the magnitude of this difference was small (VOI&lt;0.04). Conformity between modalities (CT and MRI) was in agreement for both prone and supine, DSC=0.75. Prone GBT volumes were larger than supine for both MRI and CT.Conclusion:MRI improves the extent of GBT delineation. The role of MRI guided, GBT‐targeted radiotherapy requires investigation in a clinical trial.This work was supported by a grant number APP1033237 from Cancer Australia and the National Breast Cancer Foundation.

Deformable medical image registration of pleural cavity for photodynamic therapy by using finite-element based method.

When the pleural cavity is opened during the surgery portion of pleural photodynamic therapy (PDT) of malignant mesothelioma, the pleural volume will deform. This impacts the delivered dose when using highly conformal treatment techniques. To track the anatomical changes and contour the lung and chest cavity, an infrared camera-based navigation system (NDI) is used during PDT. In the same patient, a series of computed tomography (CT) scans of the lungs are also acquired before the surgery. The reconstructed three-dimensional contours from both NDI and CTs are imported into COMSOL Multiphysics software, where a finite element-based (FEM) deformable image registration is obtained. The CT contour is registered to the corresponding NDI contour by overlapping the center of masses and aligning their orientations. The NDI contour is considered as the reference contour, and the CT contour is used as the target one, which will be deformed. Deformed Geometry model is applied in COMSOL to obtain a deformed target contour. The distortion of the volume at X, Y and Z is mapped to illustrate the transformation of the target contour. The initial assessment shows that FEM-based image deformable registration can fuse images acquired by different modalities. It provides insights into the deformation of anatomical structures along X, Y and Z-axes. The deformed contour has good matches to the reference contour after the dynamic matching process. The resulting three-dimensional deformation map can be used to obtain the locations of other critical anatomic structures, e.g., heart, during surgery.

COMPUTED TOMOGRAPHIC APPEARANCE OF THE TEMPOROMANDIBULAR JOINT IN 1018 ASYMPTOMATIC HORSES: A MULTI-INSTITUTION STUDY.

Published descriptions of nonseptic arthritis of the equine temporomandibular joint (TMJ) are rare and large studies investigating variations in the TMJ for asymptomatic horses are lacking. The objectives of this cross-sectional, retrospective, multi-institutional study were to describe anatomical variations in the TMJ detected using computed tomography (CT) in an equid population asymptomatic for TMJ disease and determine whether these variations were associated with patient signalment, reason for CT examination, or CT slice width. Medical records at eight hospitals were searched for horses that had head/neck CT scans and no clinical signs of TMJ disease. Age, breed, sex, clinical presentation, and CT slice width data were recorded. Alterations in CT contour and density of the mandibular condyles, mandibular fossae, and TMJ intra-articular discs were described for each horse. Generalized logistic regression was used to test associations between anatomical variations and horse age. A total of 1018 horses were sampled. Anatomical variations were found in TMJ CT images for 40% of horses and 29% of joints. These were dichotomous with regard to age. Horses <1 year old commonly had alterations in the shape and density of the mandibular condyle. Older horses commonly had spherical hypodensities within the mandibular condyles consistent with bone cysts; and hyperdense regions of the intra-articular disc consistent with dystrophic mineralization. Findings indicated that TMJ anatomic variations were common in CT images of younger and older horses asymptomatic for TMJ disease. Future studies are needed to more definitively characterize these CT variations using gross pathology and histopathology.

Automatic Substitute Computed Tomography Generation and Contouring for Magnetic Resonance Imaging (MRI)-Alone External Beam Radiation Therapy From Standard MRI Sequences.

To validate automatic substitute computed tomography CT (sCT) scans generated from standard T2-weighted (T2w) magnetic resonance (MR) pelvic scans for MR-Sim prostate treatment planning. A Siemens Skyra 3T MR imaging (MRI) scanner with laser bridge, flat couch, and pelvic coil mounts was used to scan 39 patients scheduled for external beam radiation therapy for localized prostate cancer. For sCT generation a whole-pelvis MRI scan (1.6 mm 3-dimensional isotropic T2w SPACE [Sampling Perfection with Application optimized Contrasts using different flip angle Evolution] sequence) was acquired. Three additional small field of view scans were acquired: T2w, T2*w, and T1w flip angle 80° for gold fiducials. Patients received a routine planning CT scan. Manual contouring of the prostate, rectum, bladder, and bones was performed independently on the CT and MR scans. Three experienced observers contoured each organ on MRI, allowing interobserver quantification. To generate a training database, each patient CT scan was coregistered to their whole-pelvis T2w using symmetric rigid registration and structure-guided deformable registration. A new multi-atlas local weighted voting method was used to generate automatic contours and sCT results. The mean error in Hounsfield units between the sCT and corresponding patient CT (within the body contour) was 0.6 ± 14.7 (mean ± 1 SD), with a mean absolute error of 40.5 ± 8.2 Hounsfield units. Automatic contouring results were very close to the expert interobserver level (Dice similarity coefficient): prostate 0.80 ± 0.08, bladder 0.86 ± 0.12, rectum 0.84 ± 0.06, bones 0.91 ± 0.03, and body 1.00 ± 0.003. The change in monitor units between the sCT-based plans relative to the gold standard CT plan for the same dose prescription was found to be 0.3% ± 0.8%. The 3-dimensional γ pass rate was 1.00 ± 0.00 (2 mm/2%). The MR-Sim setup and automatic sCT generation methods using standard MR sequences generates realistic contours and electron densities for prostate cancer radiation therapy dose planning and digitally reconstructed radiograph generation.

Deformable image registration of CT images for automatic contour propagation in radiation therapy.

Radiotherapy treatment plan may be replanned due the changes of tumors and organs at risk (OARs) during the treatment. Deformable image registration (DIR) based Computed Tomography (CT) contour propagation in the routine clinical setting is expected to reduce time needed for necessary manual tumors and OARs delineations and increase the efficiency of replanning. In this study, a DIR method was developed for CT contour propagation. Prior structure delineations were incorporated into Demons DIR, which was represented by adding an intensity matching term of the delineated tissues pairs to the energy function of Demons. The performance of our DIR was evaluated with five clinical head-and-neck and five lung cancer cases. The experimental results verified the improved accuracy of the proposed registration method compared with conventional registration and Demons DIR.

SU-E-J-111: Finite Element-Based Deformable Image Registration of Pleural Cavity for Photodynamic Therapy

Purpose: The pleural volumes will deform during surgery portion of the pleural photodynamic therapy (PDT) of lung cancer when the pleural cavity is opened. This impact the delivered dose when using highly conformal treatment techniques. In this study, a finite element-based (FEM) deformable image registration is used to quantify the anatomical variation between the contours for the pleural cavities obtained in the operating room and those determined from pre-surgery computed tomography (CT) scans. Methods: An infrared camera-based navigation system (NDI) is used during PDT to track the anatomical changes and contour the lung and chest cavity. A series of CTs of the lungs, in the same patient, are also acquired before the surgery. The structure contour of lung and the CTs are processed and contoured in Matlab and MeshLab. Then, the contours are imported into COMSOL Multiphysics 5.0, where the FEM-based deformable image registration is obtained using the deformed mesh - moving mesh (ALE) model. The NDI acquired lung contour is considered as the reference contour, and the CT contour is used as the target one, which will be deformed. Results: The reconstructed three-dimensional contours from both NDI and CT can be converted to COMSOL so that a three-dimensional ALE model can be developed. The contours can be registered using COMSOL ALE moving mesh model, which takes into account the deformation along x, y and z-axes. The deformed contour has good matches to the reference contour after the dynamic matching process. The resulting 3D deformation map can be used to obtain the locations of other critical anatomic structures, e.g., heart, during surgery. Conclusion: Deformable image registration can fuse images acquired by different modalities. It provides insights into the development of phenomenon and variation in normal anatomical structures over time. The initial assessments of three-dimensional registration show good agreement.

Comparison and Consensus Guidelines for Delineation of Clinical Target Volume for CT- and MR-Based Brachytherapy in Locally Advanced Cervical Cancer

To create and compare consensus clinical target volume (CTV) contours for computed tomography (CT) and 3-Tesla (3-T) magnetic resonance (MR) image-based cervical-cancer brachytherapy. Twenty-three experts in gynecologic radiation oncology contoured the same 3 cervical cancer brachytherapy cases: 1 stage IIB near-complete response (CR) case with a tandem and ovoid, 1 stage IIB partial response (PR) case with tandem and ovoid with needles, and 1 stage IB2 CR case with a tandem and ring applicator. The CT contours were completed before the MRI contours. These were analyzed for consistency and clarity of target delineation using an expectation maximization algorithm for simultaneous truth and performance level estimation (STAPLE), with κ statistics as a measure of agreement between participants. The conformity index was calculated for each of the 6 data sets. Dice coefficients were generated to compare the CT and MR contours of the same case. For all 3 cases, the mean tumor volume was smaller on MR than on CT (P<.001). The κ and conformity index estimates were slightly higher for CT, indicating a higher level of agreement on CT. The Dice coefficients were 89% for the stage IB2 case with a CR, 74% for the stage IIB case with a PR, and 57% for the stage IIB case with a CR. In a comparison of MR-contoured with CT-contoured CTV volumes, the higher level of agreement on CT may be due to the more distinct contrast medium visible on the images at the time of brachytherapy. MR at the time of brachytherapy may be of greatest benefit in patients with large tumors with parametrial extension that have a partial or complete response to external beam. On the basis of these results, a 95% consensus volume was generated for CT and for MR. Online contouring atlases are available for instruction at http://www.nrgoncology.org/Resources/ContouringAtlases/GYNCervicalBrachytherapy.aspx.

A comparison of rigid registration methods for prostate localization on CBCT and the dependence on rectum distension

We evaluated automatic three-dimensional intensity-based rigid registration (RR) methods for prostate localization on CBCT scans and studied the impact of rectum distension on registration quality. 106 CBCT scans of 9 prostate patients were used. Each one was registered to the planning computed tomography (CT) scan using different methods: (a) global registration, (b) pelvis bony structure registration, (c) bony registration refined by a local prostate registration using the CT clinical target volume (CTV) expanded with 1, 3, 5, 8, 10, 12, 15 or 20-mm margin. Automatic CBCT contours were generated after propagation of the manual CT contours. To evaluate results, a radiation oncologist was asked to manually delineate the CTV on the CBCT scans (gold standard). The Dice similarity coefficients between propagated and manual CBCT contours were calculated.

A study to assess the feasibility of using CT (±diagnostic MRI) instead of MRI at brachytherapy in image guided brachytherapy in cervical cancer

AbstractPurposeTo compare the contours and dose volume histograms (DVH) parameters of the high-risk clinical target volume (HRCTV) contoured on computed tomography (CT) using clinical findings at brachytherapy, clinical findings at brachytherapy with magnetic resonance imaging (MRI) at diagnosis and HRCTV defined on MRI at brachytherapy in cervical cancer patients.Materials and methodsFifteen patients undergoing MRI-guided image-based brachytherapy underwent both CT and MRI after applicator insertion. Two sets of contours were defined on CT. In the first set, the HRCTV was defined with the help of clinical findings at brachytherapy (CT-HRCTV). In the second set, HRCTV was defined with MRI at diagnosis and clinical findings at brachytherapy (CT-HRCTVdmri). This was compared with the HRCTV defined on MRI at brachytherapy (MR-HRCTV). The doses to the organs at risk (OARs) were compared for CT and MRI.ResultsA significant overestimation of the maximum width and width at point A was observed for CT-HRCTV (p −0·00; 0·00) and CT-HRCTVdmri (p −0·03; 0·01), respectively. The height was underestimated with CT-HRCTV in patients with intrauterine disease extension. For a single fraction, the mean difference in the D90 for the CT contours was &lt;1 Gy. The doses to the OARs were comparable.ConclusionsCT may be an alternative when facilities for MRI image-based brachytherapy are lacking, provided at least one MRI is available before brachytherapy.

Improved dosimetry in prostate brachytherapy using high resolution contrast enhanced magnetic resonance imaging: a feasibility study.

PurposeTo assess detailed dosimetry data for prostate and clinical relevant intra- and peri-prostatic structures including neurovascular bundles (NVB), urethra, and penile bulb (PB) from postbrachytherapy computed tomography (CT) versus high resolution contrast enhanced magnetic resonance imaging (HR-CEMRI).Material and methodsEleven postbrachytherapy prostate cancer patients underwent HR-CEMRI and CT imaging. Computed tomography and HR-CEMRI images were randomized and 2 independent expert readers created contours of prostate, intra- and peri-prostatic structures on each CT and HR-CEMRI scan for all 11 patients. Dosimetry data including V100, D90, and D100 was calculated from these contours.ResultsMean V100 values from CT and HR-CEMRI contours were as follows: prostate (98.5% and 96.2%, p = 0.003), urethra (81.0% and 88.7%, p = 0.027), anterior rectal wall (ARW) (8.9% and 2.8%, p < 0.001), left NVB (77.9% and 51.5%, p = 0.002), right NVB (69.2% and 43.1%, p = 0.001), and PB (0.09% and 11.4%, p = 0.005). Mean D90 (Gy) derived from CT and HR-CEMRI contours were: prostate (167.6 and 150.3, p = 0.012), urethra (81.6 and 109.4, p = 0.041), ARW (2.5 and 0.11, p = 0.003), left NVB (98.2 and 58.6, p = 0.001), right NVB (87.5 and 55.5, p = 0.001), and PB (11.2 and 12.4, p = 0.554).ConclusionsFindings of this study suggest that HR-CEMRI facilitates accurate and meaningful dosimetric assessment of prostate and clinically relevant structures, which is not possible with CT. Significant differences were seen between CT and HR-CEMRI, with volume overestimation of CT derived contours compared to HR-CEMRI.