Dataset Information

Automated contouring and statistical process control for plan quality in a breast clinical trial.

ABSTRACT:

Background and purpose

Automatic review of breast plan quality for clinical trials is time-consuming and has some unique challenges due to the lack of target contours for some planning techniques. We propose using an auto-contouring model and statistical process control to independently assess planning consistency in retrospective data from a breast radiotherapy clinical trial.

Materials and methods

A deep learning auto-contouring model was created and tested quantitatively and qualitatively on 104 post-lumpectomy patients' computed tomography images (nnUNet; train/test: 80/20). The auto-contouring model was then applied to 127 patients enrolled in a clinical trial. Statistical process control was used to assess the consistency of the mean dose to auto-contours between plans and treatment modalities by setting control limits within three standard deviations of the data's mean. Two physicians reviewed plans outside the limits for possible planning inconsistencies.

Results

Mean Dice similarity coefficients comparing manual and auto-contours was above 0.7 for breast clinical target volume, supraclavicular and internal mammary nodes. Two radiation oncologists scored 95% of contours as clinically acceptable. The mean dose in the clinical trial plans was more variable for lymph node auto-contours than for breast, with a narrower distribution for volumetric modulated arc therapy than for 3D conformal treatment, requiring distinct control limits. Five plans (5%) were flagged and reviewed by physicians: one required editing, two had clinically acceptable variations in planning, and two had poor auto-contouring.

Conclusions

An automated contouring model in a statistical process control framework was appropriate for assessing planning consistency in a breast radiotherapy clinical trial.

SUBMITTER: Baroudi H

PROVIDER: S-EPMC10498301 | biostudies-literature | 2023 Oct

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Automated contouring and statistical process control for plan quality in a breast clinical trial.

Baroudi Hana H Huy Minh Nguyen Callistus I CI Maroongroge Sean S Smith Benjamin D BD Niedzielski Joshua S JS Shaitelman Simona F SF Melancon Adam A Shete Sanjay S Whitaker Thomas J TJ Mitchell Melissa P MP Yvonne Arzu Isidora I Duryea Jack J Hernandez Soleil S El Basha Daniel D Mumme Raymond R Netherton Tucker T Hoffman Karen K Court Laurence L

Physics and imaging in radiation oncology 20230823

<h4>Background and purpose</h4>Automatic review of breast plan quality for clinical trials is time-consuming and has some unique challenges due to the lack of target contours for some planning techniques. We propose using an auto-contouring model and statistical process control to independently assess planning consistency in retrospective data from a breast radiotherapy clinical trial.<h4>Materials and methods</h4>A deep learning auto-contouring model was created and tested quantitatively and qu ...[more]

PMID: 37712064

Similar Datasets

Project description:BackgroundMajor Outcomes with Personalized Dialysate TEMPerature (MyTEMP) is a 4-year cluster-randomized clinical trial comparing the effect of using a personalized, temperature-reduced dialysate protocol versus a dialysate temperature of 36.5°C on cardiovascular-related death and hospitalization. Randomization was performed at the level of the dialysis center ("the cluster").ObjectiveThe objective is to outline the statistical analysis plan for the MyTEMP trial.DesignMyTEMP is a pragmatic, 2-arm, parallel-group, registry-based, open-label, cluster-randomized trial.SettingA total of 84 dialysis centers in Ontario, Canada.PatientsApproximately 13 500 patients will have received in-center hemodialysis at the 84 participating dialysis centers during the trial period (April 3, 2017, to March 1, 2021, with a maximum follow-up to March 31, 2021).MethodsPatient identification, baseline characteristics, and study outcomes will be obtained primarily through Ontario administrative health care databases held at ICES. Covariate-constrained randomization was used to allocate the 84 dialysis centers (1:1) to the intervention group or the control group. Centers in the intervention group used a personalized, temperature-reduced dialysate protocol, and centers in the control group used a fixed dialysate temperature of 36.5°C.OutcomesThe primary outcome is a composite of cardiovascular-related death or major cardiovascular-related hospitalization (defined as a hospital admission with myocardial infarction, congestive heart failure, or ischemic stroke) recorded in administrative health care databases. The key secondary outcome is the mean drop in intradialytic systolic blood pressure, defined as the patients' predialysis systolic blood pressure minus their nadir systolic blood pressure during the dialysis treatment. Anonymized data on patients' predialysis and intradialytic systolic blood pressure were collected at monthly intervals from each dialysis center.Analysis planThe primary analysis will follow an intent-to-treat approach. The primary outcome will be analyzed at the patient level as the hazard ratio of time-to-first event, estimated from a subdistribution hazards model. Within-center correlation will be accounted for using a robust sandwich estimator. In the primary analysis, patients' observation time will end if they experience the primary outcome, emigrate from Ontario, or die of a noncardiovascular cause (which will be treated as a competing risk event). The between-group difference in the mean drop in intradialytic systolic blood pressure obtained during the dialysis sessions throughout the trial period will be analyzed at the center level using an unadjusted random-effects linear mixed model.Trial statusThe MyTEMP trial period is April 3, 2017, to March 31, 2021. We expect to analyze and report results by 2023 once the updated data are available at ICES.Trial registrationMyTEMP is registered with the US National Institutes of Health at clincaltrials.gov (NCT02628366).Statistical analytic planVersion 1.1 June 15, 2021.

Project description:BackgroundArtificial intelligence (AI) shows great potential to streamline the treatment planning process. However, its clinical adoption is slow due to the limited number of clinical evaluation studies and because often, the translation of the predicted dose distribution to a deliverable plan is lacking. This study evaluates two different, deliverable AI plans in terms of their clinical acceptability based on quantitative parameters and qualitative evaluation by four radiation oncologists.MethodsFor 20 left-sided node-negative breast cancer patients, treated with a prescribed dose of 40.05 Gy, using tangential beam intensity modulated radiotherapy, two model-based treatment plans were evaluated against the corresponding manual plan. The two models used were an in-house developed U-net model and a vendor-developed contextual atlas regression forest model (cARF). Radiation oncologists evaluated the clinical acceptability of each blinded plan and ranked plans according to preference. Furthermore, a comparison with the manual plan was made based on dose volume histogram parameters, clinical evaluation criteria and preparation time.ResultsThe U-net model resulted in a higher average and maximum dose to the PTV (median difference 0.37 Gy and 0.47 Gy respectively) and a slightly higher mean heart dose (MHD) (0.01 Gy). The cARF model led to higher average and maximum doses to the PTV (0.30 and 0.39 Gy respectively) and a slightly higher MHD (0.02 Gy) and mean lung dose (MLD, 0.04 Gy). The maximum MHD/MLD difference was ≤ 0.5 Gy for both AI plans. Regardless of these dose differences, 90-95% of the AI plans were considered clinically acceptable versus 90% of the manual plans. Preferences varied between the radiation oncologists. Plan preparation time was comparable between the U-net model and the manual plan (287 s vs 253 s) while the cARF model took longer (471 s). When only considering user interaction, plan generation time was 121 s for the cARF model and 137 s for the U-net model.ConclusionsTwo AI models were used to generate deliverable plans for breast cancer patients, in a time-efficient manner, requiring minimal user interaction. Although the AI plans resulted in slightly higher doses overall, radiation oncologists considered 90-95% of the AI plans clinically acceptable.

Project description:BackgroundA key decision in the treatment of atrial fibrillation is choosing between a rhythm control strategy or a rate control strategy as the main strategy. When choosing rate control, the optimal heart rate target is uncertain. The Danish Atrial Fibrillation trial is a randomized, multicenter, two-group, superiority trial comparing strict rate control versus lenient rate control in patients with either persistent or permanent atrial fibrillation at inclusion. To prevent bias arising from selective reporting and data-driven analyses, we developed a predefined description of the statistical analysis.MethodsThe primary outcome of this trial is the physical component score of the SF-36 questionnaire. A total of 350 participants will be enrolled based on a minimal important difference of 3 points on the physical component score of the SF-36 questionnaire, a standard deviation of 10 points, a statistical power of 80% (beta of 20%), and an acceptable risk of type I error of 5%. All secondary, exploratory, and echocardiographic outcomes will be hypothesis-generating. The analyses of all outcomes will be based on the intention-to-treat principle. We will analyze continuous outcomes using linear regression adjusting for "site," type of atrial fibrillation at inclusion (persistent/ permanent), left ventricular ejection fraction (≥ 40% or < 40%), and the baseline value of the outcome (all as fixed effects). We define our threshold for statistical significance as a p-value of 0.05 and assessments of clinical significance will be based on the anticipated intervention effects defined in the sample size and power estimations. Thresholds for both statistical and clinical significance will be assessed according to the 5-step procedure proposed by Jakobsen and colleagues.DiscussionThis statistical analysis plan will be published prior to enrolment completion and before any data are available and is sought to increase the validity of the DANish Atrial Fibrillation trial.Trial registrationClinicaltrials.gov NCT04542785. Registered on Sept 09, 2020.

Project description:PurposeTo develop a new approach for interobserver variability analysis.Methods and materialsEight radiation oncologists specializing in breast cancer radiation therapy delineated a patient's left breast "from scratch" and from a template that was generated using deformable image registration. Three of the radiation oncologists had previously received training in Radiation Therapy Oncology Group consensus contouring for breast cancer atlas. The simultaneous truth and performance level estimation algorithm was applied to the 8 contours delineated "from scratch" to produce a group consensus contour. Individual Jaccard scores were fitted to a beta distribution model. We also applied this analysis to 2 or more patients, which were contoured by 9 breast radiation oncologists from 8 institutions.ResultsThe beta distribution model had a mean of 86.2%, standard deviation (SD) of ±5.9%, a skewness of -0.7, and excess kurtosis of 0.55, exemplifying broad interobserver variability. The 3 RTOG-trained physicians had higher agreement scores than average, indicating that their contours were close to the group consensus contour. One physician had high sensitivity but lower specificity than the others, which implies that this physician tended to contour a structure larger than those of the others. Two other physicians had low sensitivity but specificity similar to the others, which implies that they tended to contour a structure smaller than the others. With this information, they could adjust their contouring practice to be more consistent with others if desired. When contouring from the template, the beta distribution model had a mean of 92.3%, SD ± 3.4%, skewness of -0.79, and excess kurtosis of 0.83, which indicated a much better consistency among individual contours. Similar results were obtained for the analysis of 2 additional patients.ConclusionsThe proposed statistical approach was able to measure interobserver variability quantitatively and to identify individuals who tended to contour differently from the others. The information could be useful as feedback to improve contouring consistency.

Project description:PurposeCreate a comprehensive automated solution for pediatric and adult VMAT-CSI including contouring, planning, and plan check to reduce planning time and improve plan quality.MethodsSeventy-seven previously treated CSI patients (age, 2-67 years) were used for creation of an auto-contouring model to segment 25 organs at risk (OARs). The auto-contoured OARs were evaluated using the Dice Similarity Coefficient (DSC), 95% Hausdorff Distance (HD95), and a qualitative ranking by one physician and one physicist (scale: 1-acceptable, 2-minor edits, 3-major edits). The auto-planning script was developed using the Varian Eclipse Scripting API and tested with 20 patients previously treated with either low-dose VMAT-CSI (12 Gy) or high-dose VMAT-CSI (36 Gy + 18 Gy boost). Clinically relevant metrics, planning time, and blinded physician review were used to evaluate significance of differences between the auto and manual plans. Finally, the plan preparation for treatment and plan check processes were automated to improve efficiency and safety of VMAT-CSI.ResultsThe auto-contours achieved an average DSC of 0.71 ± 0.15, HD95 of 4.81 ± 4.68, and reviewers' ranking of 1.22 ± 0.39, indicating close to "acceptable-as-is" contours. Compared to the manual CSI plans, the auto-plans for both dose regimens achieved statistically significant reductions in body V50% and Dmean for parotids, submandibular, and thyroid glands. The variance in the dosimetric parameters decreased for the auto-plans as compared to the manual plans indicating better plan consistency. From the blinded review, the auto-plans were marked as equivalent or superior to the manual-plans 88.3% of the time. The required time for the auto-contouring and planning was consistently between 1-2 hours compared to an estimated 5-6 hours for manual contouring and planning.ConclusionsReductions in contouring and planning time without sacrificing plan quality were obtained using the developed auto-planning process. The auto-planning scripts and documentation will be made freely available to other institutions and clinics.

Project description:BackgroundTransforaminal epidural injections with steroids (TESI) are used increasingly for patients with sciatica. However, their safety, effectiveness, and cost-effectiveness are still a matter of debate. This a priori statistical analysis plan describes the methodology of the analysis for the STAR trial that assesses the (cost-)effectiveness of TESI during the acute stage of sciatica (< 8 weeks).MethodsThe STAR trial is a multicentre, randomized controlled, prospective trial (RCT) investigating the (cost-)effectiveness of TESI by making a three-group comparison among patients with acute sciatica due to a herniated lumbar disc (< 8 weeks): (1) TESI combined with levobupivacaine added to oral pain medication (intervention group 1) versus oral pain medication alone (control group), (2) intervention group 1 versus transforaminal epidural injection with levobupivacaine and saline solution added to oral pain medication (intervention group 2), and (3) intervention group 2 versus control group. Co-primary outcomes were physical functioning (Roland Morris Disability Questionnaire), pain intensity (10-point numerical rating scale), and global perceived recovery (7-point Likert scale, dichotomized into 'recovered' and 'not recovered'). For all three comparisons, we defined the following minimal clinically relevant between-group differences: two points for pain intensity (range 0-10), four points for physical functioning (range 0-24) and a 20% difference in recovery rate. Secondary outcomes are health-related quality of life (EQ-5D-5L) and patient satisfaction (7-point Likert scale) and surgery rate. We also collected resource use data to perform an economic evaluation. Analyses will be conducted by intention-to-treat with p < 0.05 (two-tailed) for all three comparisons. Effects will be estimated using mixed models by maximum likelihood. For each comparison, mean differences, or difference in proportions, between groups will be tested per time point and an overall mean difference, or difference in proportions, between groups during the complete duration of follow-up (6 months) will be estimated. In the economic evaluation, Multivariate Imputation by Chained Equations will be used to handle missing data. Cost and effect differences will be estimated using seemingly unrelated regression, and uncertainty will be estimated using bootstrapping techniques.DiscussionThis statistical analysis plan provides detailed information on the intended analysis of the STAR trial, which aims to deliver evidence about the (cost-)effectiveness of TESI during the acute phase of sciatica (< 8 weeks).Trial registrationDutch National trial register NTR4457 (6 March 2014).

Dataset Information

Automated contouring and statistical process control for plan quality in a breast clinical trial.

Background and purpose

Materials and methods

Results

Conclusions

Publications

Automated contouring and statistical process control for plan quality in a breast clinical trial.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets