Search

Background: Structure delineation is a necessary, yet time-consuming manual procedure in radiotherapy. Recently, convolutional neural networks have been proposed to speed-up and automatise this procedure, obtaining promising results. With the advent of magnetic resonance imaging (MRI)-guided radiotherapy, MR-based segmentation is becoming increasingly relevant. However, the majority of the studies investigated automatic contouring based on computed tomography (CT)., Purpose: In this study, we investigate the feasibility of clinical use of deep learning-based automatic OARs delineation on MRI., Materials and Methods: We included 150 patients diagnosed with prostate cancer who underwent MR-only radiotherapy. A three-dimensional (3D) T1-weighted dual spoiled gradient-recalled echo sequence was acquired with 3T MRI for the generation of the synthetic-CT. The first 48 patients were included in a feasibility study training two 3D convolutional networks called DeepMedic and dense V-net (dV-net) to segment bladder, rectum and femurs. A research version of an atlas-based software was considered for comparison. Dice similarity coefficient, 95% Hausdorff distances (HD 95 ), and mean distances were calculated against clinical delineations. For eight patients, an expert RTT scored the quality of the contouring for all the three methods. A choice among the three approaches was made, and the chosen approach was retrained on 97 patients and implemented for automatic use in the clinical workflow. For the successive 53 patients, Dice, HD 95 and mean distances were calculated against the clinically used delineations., Results: DeepMedic, dV-net and the atlas-based software generated contours in 60 s, 4 s and 10-15 min, respectively. Performances were higher for both the networks compared to the atlas-based software. The qualitative analysis demonstrated that delineation from DeepMedic required fewer adaptations, followed by dV-net and the atlas-based software. DeepMedic was clinically implemented. After retraining DeepMedic and testing on the successive patients, the performances slightly improved., Conclusion: High conformality for OARs delineation was achieved with two in-house trained networks, obtaining a significant speed-up of the delineation procedure. Comparison of different approaches has been performed leading to the succesful adoption of one of the neural networks, DeepMedic, in the clinical workflow. DeepMedic maintained in a clinical setting the accuracy obtained in the feasibility study.

Purpose: Accurate 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 Materials: Twelve 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., Results: Six 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., Conclusions: Using 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., (Copyright © 2017 American Society for Radiation Oncology. Published by Elsevier Inc. All rights reserved.)

Background: Accurate tumor bed delineation after breast-conserving surgery is important. However, consistency among observers on standard postoperative radiotherapy planning CT is low and volumes can be large due to seroma formation. A preoperative delineation of the tumor might be more consistent. Therefore, the purpose of this study was to determine the consistency of preoperative target volume delineation on CT and MRI for breast-conserving radiotherapy., Methods: Tumors were delineated on preoperative contrast-enhanced (CE) CT and newly developed 3D CE-MR images, by four breast radiation oncologists. Clinical target volumes (CTVs) were created by addition of a 1.5 cm margin around the tumor, excluding skin and chest wall. Consistency in target volume delineation was expressed by the interobserver variability. Therefore, the conformity index (CI), center of mass distance (dCOM) and volumes were calculated. Tumor characteristics on CT and MRI were scored by an experienced breast radiologist., Results: Preoperative tumor delineation resulted in a high interobserver agreement with a high median CI for the CTV, for both CT (0.80) and MRI (0.84). The tumor was missed on CT in 2/14 patients (14%). Leaving these 2 patients out of the analysis, CI was higher on MRI compared to CT for the GTV (p<0.001) while not for the CTV (CT (0.82) versus MRI (0.84), p=0.123). The dCOM did not differ between CT and MRI. The median CTV was 48 cm3 (range 28-137 cm3) on CT and 59 cm3 (range 30-153 cm3) on MRI (p<0.001). Tumor shapes and margins were rated as more irregular and spiculated on CE-MRI., Conclusions: This study showed that preoperative target volume delineation resulted in small target volumes with a high consistency among observers. MRI appeared to be necessary for tumor detection and the visualization of irregularities and spiculations. Regarding the tumor delineation itself, no clinically relevant differences in interobserver variability were observed. These results will be used to study the potential for future MRI-guided and neoadjuvant radiotherapy., Trial Registration: International Clinical Trials Registry Platform NTR3198.

Purpose: To evaluate and compare the maximum temperature (T(max) ) in the head after exposure to a 300 MHz radiofrequency (RF) field induced by a magnetic resonance imaging (MRI) coil using two thermal simulation methods: Pennes' bioheat equation (PBHE) and discrete vasculature (DIVA)., Materials and Methods: The electromagnetic field induced in the head by a 7T birdcage coil was simulated using finite-difference time-domain (FDTD) and validated by MRI. The specific absorption rate (SAR) distributions normalized to the 10-gram maximum or the whole-head average were used for PBHE and DIVA simulations., Results: For all cases, the T(max) in PBHE was slightly higher than in DIVA. The T(max) was 37.9-38.4°C, depending on the simulation method or perfusion rate., Conclusion: In some situations, RF exposure limited to SAR(max,10g) led to a T(max) higher than allowed by International Electrotechnical Commission (IEC) regulations. Therefore, it is advisable to use thermal simulations to evaluate RF safety of MRI. The simulation method used only slightly influenced the observed maximum temperature; the observed temperature with PBHE was higher in all situations. So PBHE is an appropriate method for RF safety assessment of MRI in the head. Using DIVA simulations, it was found unlikely that the body temperature increases significantly due to energy deposited by a head coil under normal circumstances., (Copyright © 2011 Wiley Periodicals, Inc.)

In a multi-atlas based segmentation procedure, propagated atlas segmentations must be combined in a label fusion process. Some current methods deal with this problem by using atlas selection to construct an atlas set either prior to or after registration. Other methods estimate the performance of propagated segmentations and use this performance as a weight in the label fusion process. This paper proposes a selective and iterative method for performance level estimation (SIMPLE), which combines both strategies in an iterative procedure. In subsequent iterations the method refines both the estimated performance and the set of selected atlases. For a dataset of 100 MR images of prostate cancer patients, we show that the results of SIMPLE are significantly better than those of several existing methods, including the STAPLE method and variants of weighted majority voting.

For breast boost radiotherapy or accelerated partial breast irradiation, the tumor bed (TB) is delineated by the radiation oncologist on a planning computed tomography (CT) scan. The aim of the present study was to investigate whether the interobserver variability (IOV) of the TB delineation is reduced by providing the radiation oncologist with additional magnetic resonance imaging (MRI) or CT scans. A total of 14 T1-T2 breast cancer patients underwent a standard planning CT in the supine treatment position following lumpectomy, as well as additional pre- and postoperative imaging in the same position. Post-lumpectomy TBs were independently delineated by four breast radiation oncologists on standard postoperative CT and on CT registered to an additional imaging modality. The additional imaging modalities used were postoperative MRI, preoperative contrast-enhanced (CE)-CT and preoperative CE-MRI. A cavity visualization score (CVS) was assigned to each standard postoperative CT by each observer. In addition, the conformity index (CI), volume and distance between centers of mass (dCOM) of the TB delineations were calculated. On CT, the median CI was 0.57, with a median volume of 22 cm3 and dCOM of 5.1 mm. The addition of postoperative MRI increased the median TB volume significantly to 28 cm3 (P<0.001), while the CI (P=0.176) and dCOM (P=0.110) were not affected. The addition of preoperative CT or MRI increased the TB volume to 26 and 25 cm3, respectively (both P<0.001), while the CI increased to 0.58 and 0.59 (both P<0.001) and the dCOM decreased to 4.7 mm (P=0.004) and 4.6 mm (P=0.001), respectively. In patients with CVS=3, the median CI was 0.40 on CT, which was significantly increased by all additional imaging modalities, up to 0.52, and was accompanied by a median volume increase up to 6 cm3. In conclusion, the addition of postoperative MRI, preoperative CE-CT or preoperative CE-MRI did not result in a considerable reduction in the IOV in postoperative CT-guided TB delineation, while target volumes marginally increased. The value of additional imaging may be dependent on CVS. [ABSTRACT FROM AUTHOR]

Abstract: The delineation of tumors and their surrounding organs at risk is a critical step of the treatment planning for radiation therapy. Besides computer tomography (CT), other imaging modalities are used to improve the quality of the delineations, such as magnetic resonance imaging (MRI) and positron emission tomography (PET). A practical framework is presented for using multiple datasets from different modalities during the delineation phase. The system is based on two basic principles. First, all image datasets of all available modalities are displayed in their original form (in their own coordinate system, with their own spatial resolution and voxel aspect ratio), and second, delineations can take place on all orthogonal views of each dataset and changes made to a delineation are visualized in all image sets, giving direct feedback to the delineator. The major difference between the described approach and other existing delineation tools is that instead of resampling the image sets, the delineations are transformed from one dataset to another. The transformation used for transferring the delineations is obtained by rigid normalized mutual information registration. The crucial components and the benefits of the application are presented and discussed. [Copyright &y& Elsevier]

Purpose: In a randomized focal dose escalation radiation therapy trial for prostate cancer (FLAME), up to 95 Gy was prescribed to the tumor in the dose-escalated arm, with 77 Gy to the entire prostate in both arms. As dose constraints to organs at risk had priority over dose escalation and suboptimal planning could occur, we investigated how well the dose to the tumor was boosted. We developed an anatomy-based prediction model to identify plans with suboptimal tumor dose and performed replanning to validate our model.Methods and Materials: We derived dose-volume parameters from planned dose distributions of 539 FLAME trial patients in 4 institutions and compared them between both arms. In the dose-escalated arm, we determined overlap volume histograms and derived features representing patient anatomy. We predicted tumor D98% with a linear regression on anatomic features and performed replanning on 21 plans.Results: In the dose-escalated arm, the median tumor D50% and D98% were 93.0 and 84.7 Gy, and 99% of the tumors had a dose escalation greater than 82.4 Gy (107% of 77 Gy). In both arms organs at risk constraints were met. Five out of 73 anatomic features were found to be predictive for tumor D98%. Median predicted tumor D98% was 4.4 Gy higher than planned D98%. Upon replanning, median tumor D98% increased by 3.0 Gy. A strong correlation between predicted increase in D98% and realized increase upon replanning was found (ρ = 0.86).Conclusions: Focal dose escalation in prostate cancer was feasible with a dose escalation to 99% of the tumors. Replanning resulted in an increased tumor dose that correlated well with the prediction model. The model was able to identify tumors on which a higher boost dose could be planned. The model has potential as a quality assessment tool in focal dose escalated treatment plans. [ABSTRACT FROM AUTHOR]

Highlights • 5 patients with pelvic lymph node metastases received SBRT using a 1.5 T MR-linac. • Session time was <60 min for all 25 treatment fractions. • All quality assurance tests were passed (dose calculations & film measurements). Abstract Online adaptive radiotherapy using the 1.5 Tesla MR-linac is feasible for SBRT (5 × 7 Gy) of pelvic lymph node oligometastases. The workflow allows full online planning based on daily anatomy. Session duration is less than 60 min. Quality assurance tests, including independent 3D dose calculations and film measurements were passed. [ABSTRACT FROM AUTHOR]

Purpose To date no guidelines are available for contouring prostate cancer inside the gland, as visible on multiparametric (mp-) MRI. We assessed inter-institutional differences in interpretation of mp-MRI in the multicenter phase III FLAME trial. Methods We analyzed clinical delineations on mp-MRI and clinical characteristics from 260 patients across three institutes. We performed a logistic regression analysis to examine each institute’s weighting of T2w, ADC and K trans intensity maps in the delineation of the cancer. As reviewing of all delineations by an expert panel is not feasible, we made a selection based on discrepancies between a published tumor probability (TP) model and each institute’s clinical delineations using Areas Under the ROC Curve (AUC) analysis. Results Regression coefficients for the three institutes were −0.07, −0.27 and −0.11 for T2w, −1.96, −0.53 and −0.65 for ADC and 0.15, 0.20 and 0.62 for K trans , with significant differences between institutes for ADC and K trans . AUC analysis showed median AUC values of 0.92, 0.80 and 0.79. Five patients with lowest AUC values were reviewed by a uroradiologist. Conclusion Regression coefficients revealed considerably different interpretations of mp-MRI in tumor contouring between institutes and demonstrated the need for contouring guidelines. Based on AUC values outlying delineations could efficiently be identified for review. [ABSTRACT FROM AUTHOR]