Variability In Volume Delineation Research Articles

Image fusion technique for target volume delineation in 125I seed implant brachytherapy for parotid gland cancers.

Variability in volume delineation is a possible error source in brachytherapy. This study assessed the interobserver variations in clinical target volume (CTV) delineation in postoperative adjuvant 125I seed implant brachytherapy after parotid gland cancer surgical resection and evaluated the image fusion technique for target volume delineation. Five radiation oncologists delineated gross tumor volume (GTV) and CTV in 20 patients using conventional delineation and image fusion methods. The consistency in target volume delineation was determined on the basis of differences between the oncologists. Variability was determined using Kendall's W-test, the mean conformity index (CI), the mean distance to conformity (MDC), and the center of gravity distance (CGD). There were significant variations in the delineated target volumes among radiation oncologists, but the CTV consistency was significantly enhanced using the image fusion technique, based on Kendall's W, mean CI, average MDC, and average CGD, which were 0.752, 0.41, 2.75, and 4.997, respectively, using the conventional method, and 0.987, 0.86, 0.55, and 1.27, respectively, using the image fusion method. The interobserver variation in the delineation of the postoperative parotid target volume is large, but it can be considerably decreased using image fusion technology, which resulted in a noticeable improvement in the delineation precision of the target volume for parotid gland cancer. Thus, this technology can enhance the efficacy of 125I seed implant brachytherapy and decrease any adverse effects induced by errors in target delineation.

The impact of a radiologist-led workshop on MRI target volume delineation for radiotherapy.

IntroductionMagnetic resonance imaging (MRI) is increasingly used for target volume delineation in radiotherapy due to its superior soft tissue visualisation compared to computed tomography (CT). The aim of this study was to assess the impact of a radiologist‐led workshop on inter‐observer variability in volume delineation on MRI.MethodsData from three separate studies evaluating the impact of MRI in lung, breast and cervix were collated. At pre‐workshop evaluation, observers involved in each clinical site were instructed to delineate specified volumes. Radiologists specialising in each cancer site conducted an interactive workshop on interpretation of images and anatomy for each clinical site. At post‐workshop evaluation, observers repeated delineation a minimum of 2 weeks after the workshops. Inter‐observer variability was evaluated using dice similarity coefficient (DSC) and volume similarity (VOLSIM) index comparing reference and observer volumes.ResultsPost‐workshop primary gross tumour volumes (GTV) were smaller than pre‐workshop volumes for lung with a mean percentage reduction of 10.4%. Breast clinical target volumes (CTV) were similar but seroma volumes were smaller post‐workshop on both supine (65% reduction) and prone MRI (73% reduction). Based on DSC scores, improvement in inter‐observer variability was seen for the seroma cavity volume on prone MRI with a reduction in DSC score range from 0.4–0.8 to 0.7–0.9. Breast CTV demonstrated good inter‐observer variability scores (mean DSC 0.9) for both pre‐ and post‐workshop. Post‐workshop observer delineated cervix GTV was smaller than pre‐workshop by 26.9%.ConclusionA radiologist‐led workshop did not significantly reduce inter‐observer variability in volume delineation for the three clinical sites. However, some improvement was noted in delineation of breast CTV, seroma volumes and cervix GTV.

Impact of Interobserver Variability in Volume Delineation on Deformable Image Registration for Prostate

Impact of Interobserver Variability in Volume Delineation on Deformable Image Registration for Prostate

Inter-scan and inter-observer tumour volume delineation variability on cone beam computed tomography in patients treated with stereotactic body radiation therapy for early-stage non-small cell lung cancer.

Quantification of volume changes on cone beam computed tomography (CBCT) during lung stereotactic body radiation therapy (SBRT) for non-small cell lung cancer (NSCLC) may provide a useful radiological marker for radiation response and for adaptive treatment planning. This study quantifies inter-scan and inter-observer variability in tumour volume delineation on CBCT. Three clinicians independently contoured the primary gross tumour volume (GTV) manually on CBCTs taken immediately before SBRT treatment (pre) and after the same SBRT treatment (post) for 19 NSCLC patients. Relative volume differences (RVD) were calculated between the pre- and post-CBCTs for a given treatment and between any two of three observers for a given CBCT. Coefficient of variation (CV) was used to quantitatively measure and compare the extent of variability. Inter-observer variability had a significantly higher CV of 0.15±0.13 compared to inter-scan CV of 0.03±0.04 with P<0.0001. The greatest variability was observed with tumours (<2cm in diameter) versus larger tumours with 95% limit of agreement (LOA) (Mean±Standard Deviation) of 1.90%±19.55% vs. -0.97%±12.26% for inter-scan RVD and 29.99%± 73.84% vs. 9.37%±29.95% for inter-observer RVD respectively. Inter-observer variability was greater than inter-scan variability for tumour volume delineation on CBCT with greatest variability for small tumours (<2cm in diameter). LOA for inter-scan variability (~12%) helps defines a threshold for clinically meaningful tumour volume change during SBRT treatment for tumours with diameter greater than 2cm, with larger thresholds needed for smaller tumours.

Comparison of Automated Atlas-Based Segmentation Software for Postoperative Prostate Cancer Radiotherapy.

Automated atlas-based segmentation (ABS) algorithms present the potential to reduce the variability in volume delineation. Several vendors offer software that are mainly used for cranial, head and neck, and prostate cases. The present study will compare the contours produced by a radiation oncologist to the contours computed by different automated ABS algorithms for prostate bed cases, including femoral heads, bladder, and rectum. Contour comparison was evaluated by different metrics such as volume ratio, Dice coefficient, and Hausdorff distance. Results depended on the volume of interest showed some discrepancies between the different software. Automatic contours could be a good starting point for the delineation of organs since efficient editing tools are provided by different vendors. It should become an important help in the next few years for organ at risk delineation.

A review of interventions to reduce inter-observer variability in volume delineation in radiation oncology.

Inter-observer variability (IOV) in target volume and organ-at-risk (OAR) delineation is a source of potential error in radiation therapy treatment. The aims of this study were to identify interventions shown to reduce IOV in volume delineation. Medline and Pubmed databases were queried for relevant articles using various keywords to identify articles which evaluated IOV in target or OAR delineation for multiple (>2) observers. The search was limited to English language articles and to those published from 1 January 2000 to 31 December 2014. Reference lists of identified articles were scrutinised to identify relevant studies. Studies were included if they reported IOV in contouring before and after an intervention including the use of additional or alternative imaging. Fifty-six studies were identified. These were grouped into evaluation of guidelines (n=9), teaching (n=9), provision of an autocontour (n=7) and the impact of imaging (n=31) on IOV. Guidelines significantly reduced IOV in 7/9 studies. Teaching interventions reduced IOV in 8/9 studies, statistically significant in 4. The provision of an autocontour improved consistency of contouring in 6/7 studies, statistically significant in 5. The effect of additional imaging on IOV was variable. Pre-operative CT was useful in reducing IOV in contouring breast and liver cancers, PET scans in lung cancer, rectal cancer and lymphoma and MRI scans in OARs in head and neck cancers. Inter-observer variability in volume delineation can be reduced with the use of guidelines, provision of autocontours and teaching. The use of multimodality imaging is useful in certain tumour sites.

PO-0782 TARGET DELINEATION IN RECTAL CANCER RADIOTHERAPY: RADIOBIOLOGICAL CONSEQUENCES OF ATLAS IMPLEMENTATION

Purpose/Objective: Target volume delineation variability is a major concern for cooperative groups and multi-center clinical trials. The aim of this study is to ascertain the dosimetric and radiobiologic impact of using a consensus guideline target volume delineation atlas regarding the reduction of inter-observer target delineation variation. Materials and Methods: Using a representative case and target volume delineation instructions and a standardized case derived from a proposed SWOG IMRT rectal cancer clinical trial, gross tumor volume (GTV), and clinical target volumes (CTV) were contoured by 13 physician observers. Observers were then randomly assigned to receipt (atlas) or non-receipt (control) of a cooperative-group supported consensus guideline/atlas for anorectal cancers, then instructed to re-contour the same case. PTV margin expansion and plan optimization were performed using standardized parameters from the study protocol, using sequential dynamic arc IMRT. The calculation of tumor control probability (TCP), normal tissue complication probability (NTCP), complication-free tumor control probability (P+), biologically effective uniform dose and generalized equivalent uniform dose (gEUD) was performed for all the plans. Grouped Atlas/Control TCP values were compared using Wilcoxon rank sum test and Levene’s test of variance equality for the grouped cohort GTV, CTV, and PTV inter-observer TCP grand mean (TCP’; i.e. the mean of all interobserver TCP calculations) Results: Although the variation of P+ among the members of the atlasassisted group (0.8%) was larger than that of the control group (0.5%) initially, in the second delineation the reduction of the deviation was lower for the atlas-assisted group (0.7% against 1.0%). Atlas implementation resulted in mean±SD P+ of 85.7±0.8, 78.1±2.0, and 76.8±1.1 for GTV, CTV, and PTV respectively, while for the control group the equivalent values were 85.1±0.6, 78.1±1.3, 76.4±0.6. Reduction in TCP’ variation between plans was observed with atlas use for CTV and PTV ROIs, with a reduction in TCP’ difference between plans from 13.7% (pre-atlas) to <1% (post atlas) for CTV, and from 32.5% (pre-atlas) to 7.7% (post-atlas) [p<0.01]. For the control cohort, no statistically significant differential in grand mean difference was noted, with TCP’ values for CTV of 10.2 and 6.5%, and PTV of 23.6% and 28.2% on sequential recontouring. A <1% TCP’ difference was observed in TCP for both atlas or control cohorts, suggesting minimal influence of the atlas upon GTV definition. Conclusions: The use of a visual atlas and consensus treatment guidelines in the development of rectal cancer IMRT treatment plans resulted in at least 4-fold reduction in the inter-observer dependent radiobiologic variability in TCP’ for CTV and PTV volumes. The observed differences in tumor control probability for the CTV and PTV due to the use of the atlas are potentially clinically meaningful, and suggest routine atlas implementation is a key component for reducing potential confounders of outcome in conformal radiotherapy trials. PO-0783 CONE BEAM CT BASED MARGIN GENERATION: DO WE REQUIRE DIFFERENT MARGINS FOR SUB-REGIONS OF HEAD AND NECK? A. Budrukkar 1 , R. Krishnatry 1 , K. Rajput 2 , S. Yadadu 2 , B. George 2 , S. Chaudhari 2 , S. Ghosh Laskar 1 , V. Murthy 1 , T. Gupta 1 , J. Agarwal 1

Inter-observer and Intra-observer Target Volume Delineation Variability in Oropharyngeal Cancer Patients with Significant Dental Artifact on CT

Inter-observer and Intra-observer Target Volume Delineation Variability in Oropharyngeal Cancer Patients with Significant Dental Artifact on CT

Internal target volume defined by contrast-enhanced 4D-CT scan in unresectable pancreatic tumour: Evaluation and reproducibility

We compared customized ITVs obtained with CE-4D-CT imaging (ITV 4D) with a population-based (ITV PBC) in 29 patients (PTs) and evaluated the intra-observer ITV delineation reproducibility in 5 PTs with unresectable pancreatic ductal adenocarcinoma (PDA). The ITV PBC was quite different from the ITV 4D, with under/over estimation of volume. Intra-observer volume delineation variability on CE-4D-CT and on a single-phase CE-CT were similar (27.6% vs 24.9%).

Radiation treatment planning with an integrated positron emission and computer tomography (PET/CT): a feasibility study

Purpose To investigate the usefulness of hardware coregistered PET/CT images for target volume definition. Methods and materials Thirty-nine patients presenting with various solid tumors were investigated. CT and a FDG-PET were obtained in treatment position in an integrated PET/CT scanner, and coregistered images were used for treatment planning. First, volume delineation was performed on the CT data. In a second step, the corresponding PET data were used as an overlay to the CT data to define the target volume. Delineation was done independently by two investigators. Results Coregistered PET/CT showed good fusion accuracy. The GTV increased by 25% or more because of PET in 17% of cases with head-and-neck (2/12) and lung cancer (1/6), and in 33% (7/21) in cancer of the pelvis. The GTV was reduced ≥25% in 33% of patients with head-and-neck cancer (4/12), in 67% with lung cancer (4/6), and 19% with cancer of the pelvis (4/21). Overall, in 56% (22/39) of cases, GTV delineation was changed significantly if information from metabolic imaging was used in the planning process. The modification of the GTV translated into altered PTV changes exceeding >20% in 46% (18/39) of cases. With PET, volume delineation variability between two independent oncologists decreased from a mean volume difference of 25.7 cm 3 to 9.2 cm 3 associated with a reduction of the standard deviation from 38.3 cm 3 to 13.3 cm 3 ( p = 0.02). In 16% of cases, PET/CT revealed distant metastasies, changing the treatment strategy from curative to palliative. Conclusion Integrated PET/CT for treatment planning for three-dimensional conformal radiation therapy improves the standardization of volume delineation compared with that of CT alone. PET/CT has the potential for reducing the risk for geographic misses, to minimize the dose of ionizing radiation applied to non-target organs, and to change the current practice to three-dimensional conformal radiation therapy planning by taking into account the metabolic and biologic features of cancer. The impact on treatment outcome remains to be demonstrated.