Recommendations on how to establish evidence from auto-segmentation software in radiotherapy

Deep learning for autosegmentation for radiotherapy treatment planning: State-of-the-art and novel perspectives

The rapid development of artificial intelligence (AI) has gained importance, with many tools already entering our daily lives. The medical field of radiation oncology is also subject to this development, with AI entering all steps of the patient journey. In this review article, we summarize contemporary AI techniques and explore the clinical applications of AI-based automated segmentation models in radiotherapy planning, focusing on delineation of organs at risk (OARs), the gross tumor volume (GTV), and the clinical target volume (CTV). Emphasizing the need for precise and individualized plans, we review various commercial and freeware segmentation tools and also state-of-the-art approaches. Through our own findings and based on the literature, we demonstrate improved efficiency and consistency as well as time savings in different clinical scenarios. Despite challenges in clinical implementation such as domain shifts, the potential benefits for personalized treatment planning are substantial. The integration of mathematical tumor growth models and AI-based tumor detection further enhances the possibilities for refining target volumes. As advancements continue, the prospect of one-stop-shop segmentation and radiotherapy planning represents an exciting frontier in radiotherapy, potentially enabling fast treatment with enhanced precision and individualization.

Cardiac substructure delineation in radiation therapy - A state-of-the-art review

Delineation of cardiac substructures is crucial for a better understanding of radiation-related cardiotoxicities and to facilitate accurate and precise cardiac dose calculation for developing and applying risk models. This review examines recent advancements in cardiac substructure delineation in the radiation therapy (RT) context, aiming to provide a comprehensive overview of the current level of knowledge, challenges and future directions in this evolving field. Imaging used for RT planning presents challenges in reliably visualising cardiac anatomy. Although cardiac atlases and contouring guidelines aid in standardisation and reduction of variability, significant uncertainties remain in defining cardiac anatomy. Coupled with the inherent complexity of the heart, this necessitates auto-contouring for consistent large-scale data analysis and improved efficiency in prospective applications. Auto-contouring models, developed primarily for breast and lung cancer RT, have demonstrated performance comparable to manual contouring, marking a significant milestone in the evolution of cardiac delineation practices. Nevertheless, several key concerns require further investigation. There is an unmet need for expanding cardiac auto-contouring models to encompass a broader range of cancer sites. A shift in focus is needed from ensuring accuracy to enhancing the robustness and accessibility of auto-contouring models. Addressing these challenges is paramount for the integration of cardiac substructure delineation and associated risk models into routine clinical practice, thereby improving the safety of RT for future cancer patients.

NRG Oncology Assessment of Artificial Intelligence Deep Learning-Based Auto-segmentation for Radiation Therapy: Current Developments, Clinical Considerations, and Future Directions

Deep learning neural networks (DLNN) in Artificial intelligence (AI) have been extensively explored for automatic segmentation in radiotherapy (RT). In contrast to traditional model-based methods, data-driven AI-based models for auto-segmentation have shown high accuracy in early studies in research settings and controlled environment (single institution). Vendor-provided commercial AI models are made available as part of the integrated treatment planning system (TPS) or as a stand-alone tool that provides streamlined workflow interacting with the main TPS. These commercial tools have drawn clinics' attention thanks to their significant benefit in reducing the workload from manual contouring and shortening the duration of treatment planning. However, challenges occur when applying these commercial AI-based segmentation models to diverse clinical scenarios, particularly in uncontrolled environments. Contouring nomenclature and guideline standardization has been the main task undertaken by the NRG Oncology. AI auto-segmentation holds the potential clinical trial participants to reduce interobserver variations, nomenclature non-compliance, and contouring guideline deviations. Meanwhile, trial reviewers could use AI tools to verify contour accuracy and compliance of those submitted datasets. In recognizing the growing clinical utilization and potential of these commercial AI auto-segmentation tools, NRG Oncology has formed a working group to evaluate the clinical utilization and potential of commercial AI auto-segmentation tools. The group will assess in-house and commercially available AI models, evaluation metrics, clinical challenges, and limitations, as well as future developments in addressing these challenges. General recommendations are made in terms of the implementation of these commercial AI models, as well as precautions in recognizing the challenges and limitations.

Evaluating the relationship between contouring variability and modelled treatment outcome for prostate bed radiotherapy

Objectives.Contouring similarity metrics are often used in studies of inter-observer variation and automatic segmentation but do not provide an assessment of clinical impact. This study focused on post-prostatectomy radiotherapy and aimed to (1) identify if there is a relationship between variations in commonly used contouring similarity metrics and resulting dosimetry and (2) identify the variation in clinical target volume (CTV) contouring that significantly impacts dosimetry.Approach.The study retrospectively analysed CT scans of 10 patients from the TROG 08.03 RAVES trial. The CTV, rectum, and bladder were contoured independently by three experienced observers. Using these contours reference simultaneous truth and performance level estimation (STAPLE) volumes were established. Additional CTVs were generated using an atlas algorithm based on a single benchmark case with 42 manual contours. Volumetric-modulated arc therapy (VMAT) treatment plans were generated for the observer, atlas, and reference volumes. The dosimetry was evaluated using radiobiological metrics. Correlations between contouring similarity and dosimetry metrics were calculated using Spearman coefficient (Γ). To access impact of variations in planning target volume (PTV) margin, the STAPLE PTV was uniformly contracted and expanded, with plans created for each PTV volume. STAPLE dose-volume histograms (DVHs) were exported for plans generated based on the contracted/expanded volumes, and dose-volume metrics assessed.Mainresults. The study found no strong correlations between the considered similarity metrics and modelled outcomes. Moderate correlations (0.5 <Γ< 0.7) were observed for Dice similarity coefficient, Jaccard, and mean distance to agreement metrics and rectum toxicities. The observations of this study indicate a tendency for variations in CTV contraction/expansion below 5 mm to result in minor dosimetric impacts.Significance. Contouring similarity metrics must be used with caution when interpreting them as indicators of treatment plan variation. For post-prostatectomy VMAT patients, this work showed variations in contours with an expansion/contraction of less than 5 mm did not lead to notable dosimetric differences, this should be explored in a larger dataset to assess generalisability.

Deep learning for automated segmentation in radiotherapy: a narrative review

The segmentation of organs and structures is a critical component of radiation therapy planning, with manual segmentation being a laborious and time-consuming task. Interobserver variability can also impact the outcomes of radiation therapy. Deep neural networks have recently gained attention for their ability to automate segmentation tasks, with convolutional neural networks (CNNs) being a popular approach. This article provides a descriptive review of the literature on deep learning (DL) techniques for segmentation in radiation therapy planning. This review focuses on five clinical sub-sites and finds that U-net is the most commonly used CNN architecture. The studies using DL for image segmentation were included in brain, head and neck, lung, abdominal, and pelvic cancers. The majority of DL segmentation articles in radiation therapy planning have concentrated on normal tissue structures. N-fold cross-validation was commonly employed, without external validation. This research area is expanding quickly, and standardization of metrics and independent validation are critical to benchmarking and comparing proposed methods.

A network score-based metric to optimize the quality assurance of automatic radiotherapy target segmentations

Background and purpose

Existing methods for quality assurance of the radiotherapy auto-segmentations focus on the correlation between the average model entropy and the Dice Similarity Coefficient (DSC) only. We identified a metric directly derived from the output of the network and correlated it with clinically relevant metrics for contour accuracy.

Materials and Methods

Magnetic Resonance Imaging auto-segmentations were available for the gross tumor volume for cervical cancer brachytherapy (106 segmentations) and for the clinical target volume for rectal cancer external-beam radiotherapy (77 segmentations). The nnU-Net's output before binarization was taken as a score map. We defined a metric as the mean of the voxels in the score map above a threshold (λ). Comparisons were made with the mean and standard deviation over the score map and with the mean over the entropy map. The DSC, the 95th Hausdorff distance, the mean surface distance (MSD) and the surface DSC were computed for segmentation quality. Correlations between the studied metrics and model quality were assessed with the Pearson correlation coefficient (r). The area under the curve (AUC) was determined for detecting segmentations that require reviewing.

Results

For both tasks, our metric (λ = 0.30) correlated more strongly with the segmentation quality than the mean over the entropy map (for surface DSC, r > 0.65 vs. r < 0.60). The AUC was above 0.84 for detecting MSD values above 2 mm.

Conclusions

Our metric correlated strongly with clinically relevant segmentation metrics and detected segmentations that required reviewing, indicating its potential for automatic quality assurance of radiotherapy target auto-segmentations.

Effectiveness of bladder filling control during online MR-guided adaptive radiotherapy for rectal cancer

BACKGROUND

Magnetic resonance-guided adaptive radiotherapy (MRgART) treatment sessions at MR-Linac are time-consuming and changes in organs at risk volumes can impact the treatment dosimetry. This study aims to evaluate the feasibility to control bladder filling during the rectum MRgART online session and its effectiveness on plan dosimetry.

METHODS

A total of 109 online adaptive sessions of 24 rectum cancer patients treated at Unity 1.5 T MR-Linac with a short course radiotherapy (25 Gy, 5 Gy × 5) for whom the adaptive plan was optimized and recalculated online based on the daily magnetic resonance imaging (MRI) were analysed. Patients were fitted with a bladder catheter to control bladder filling; the bladder is emptied and then partially filled with a known amount of saline at the beginning and end of the online session. A first MRI ([Formula: see text]) acquired at the beginning of the session was used for plan adaptation and the second ([Formula: see text]) was acquired while approving the adapted plan and rigidly registered with the first to ensure the appropriateness of the isodoses on the ongoing delivery treatment. For each fraction, the time interval between the two MRIs and potential bladder changes were assessed with independent metrics, and the impact on the plan dosimetry was evaluated by comparing target and organs at risk dose volume histogram cut-off points of the plan adapted on [Formula: see text] and recalculated on [Formula: see text].

RESULTS

Median bladder volume variations, DSC, and HD of 8.17%, 0.922, and 2.92 mm were registered within a median time of 38 min between [Formula: see text] and [Formula: see text]; dosimetric differences < 0.65% were registered for target coverage, and < 0.5% for bladder, small bowel and femoral heads constraints, with a p value > 0.05.

CONCLUSION

The use of a bladder filling control procedure can help ensure the dosimetric accuracy of the online adapted treatment delivered.

A Review of the Metrics Used to Assess Auto-Contouring Systems in Radiotherapy

Auto-contouring could revolutionise future planning of radiotherapy treatment. The lack of consensus on how to assess and validate auto-contouring systems currently limits clinical use. This review formally quantifies the assessment metrics used in studies published during one calendar year and assesses the need for standardised practice. A PubMed literature search was undertaken for papers evaluating radiotherapy auto-contouring published during 2021. Papers were assessed for types of metric and the methodology used to generate ground-truth comparators. Our PubMed search identified 212 studies, of which 117 met the criteria for clinical review. Geometric assessment metrics were used in 116 of 117 studies (99.1%). This includes the Dice Similarity Coefficient used in 113 (96.6%) studies. Clinically relevant metrics, such as qualitative, dosimetric and time-saving metrics, were less frequently used in 22 (18.8%), 27 (23.1%) and 18 (15.4%) of 117 studies, respectively. There was heterogeneity within each category of metric. Over 90 different names for geometric measures were used. Methods for qualitative assessment were different in all but two papers. Variation existed in the methods used to generate radiotherapy plans for dosimetric assessment. Consideration of editing time was only given in 11 (9.4%) papers. A single manual contour as a ground-truth comparator was used in 65 (55.6%) studies. Only 31 (26.5%) studies compared auto-contours to usual inter- and/or intra-observer variation. In conclusion, significant variation exists in how research papers currently assess the accuracy of automatically generated contours. Geometric measures are the most popular, however their clinical utility is unknown. There is heterogeneity in the methods used to perform clinical assessment. Considering the different stages of system implementation may provide a framework to decide the most appropriate metrics. This analysis supports the need for a consensus on the clinical implementation of auto-contouring.

Personalized auto-segmentation for magnetic resonance imaging guided adaptive radiotherapy of prostate cancer

PURPOSE

Fast and accurate delineation of organs on treatment-fraction images is critical in magnetic resonance imaging-guided adaptive radiotherapy (MRIgART). This study proposes a personalized auto-segmentation (AS) framework to assist online delineation of prostate cancer using MRIgART.

METHODS

Image data from 26 patients diagnosed with prostate cancer and treated using hypofractionated MRIgART (5 fractions per patient) were collected retrospectively. Daily pretreatment T2-weighted MRI was performed using a 1.5-T MRI system integrated into a Unity MR-linac. First-fraction image and contour data from 16 patients (80 image-sets) were used to train the population AS model, and the remaining 10 patients composed the test set. The proposed personalized AS framework contained two main steps. First, a convolutional neural network was employed to train the population model using the training set. Second, for each test patient, the population model was progressively fine-tuned with manually checked delineations of the patient's current and previous fractions to obtain a personalized model that was applied to the next fraction.

RESULTS

Compared with the population model, the personalized models substantially improved the mean Dice similarity coefficient from 0.79 to 0.93 for the prostate clinical target volume (CTV), 0.91 to 0.97 for the bladder, 0.82 to 0.92 for the rectum, 0.91 to 0.93 for the femoral heads, respectively.

CONCLUSIONS

The proposed method can achieve accurate segmentation and potentially shorten the overall online delineation time of MRIgART. This article is protected by copyright. All rights reserved.

Real-world analysis of manual editing of deep learning contouring in the thorax region

Background and purpose

User-adjustments after deep-learning (DL) contouring in radiotherapy were evaluated to get insight in real-world editing during clinical practice. This study assessed the amount, type and spatial regions of editing of auto-contouring for organs-at-risk (OARs) in routine clinical workflow for patients in the thorax region.

Materials and methods

A total of 350 lung cancer and 362 breast cancer patients, contoured between March 2020 and March 2021 using a commercial DL-contouring method followed by manual adjustments were retrospectively analyzed. Subsampling was performed for some OARs, using an inter-slice gap of 1-3 slices. Commonly-used whole-organ contouring assessment measures were calculated, and all cases were registered to a common reference shape per OAR to identify regions of manual adjustment. Results were expressed as the median, 10th-90th percentile of adjustment and visualized using 3D renderings.

Results

Per OAR, the median amount of editing was below 1 mm. However, large adjustments were found in some locations for most OARs. In general, enlarging of the auto-contours was needed. Subsampling DL-contours showed less adjustments were made in the interpolated slices compared to simulated no-subsampling for these OARs.

Conclusion

The real-world performance of automatic DL-contouring software was evaluated and proven useful in clinical practice. Specific regions-of-adjustment were identified per OAR in the thorax region, and separate models were found to be necessary for specific clinical indications different from training data. This analysis showed the need to perform routine clinical analysis especially when procedures or acquisition protocols change to have the best configuration of the workflow.

Knowledge-based quality control of organ delineations in radiation therapy

PURPOSE

To reduce the likelihood of errors in organ delineations used for radiotherapy treatment planning, a knowledge-based quality control (KBQC) system, which discriminates between valid and anomalous delineations is developed.

METHOD AND MATERIALS

The KBQC is comprised of a group-wise inference system and anomaly detection modules trained using historical priors from 296 locally advanced lung and prostate cancer patient computational tomographies (CTs). The inference system discriminates different organs based on shape, relational, and intensity features. For a given delineated image set, the inference system solves a combinatorial optimization problem that results in an organ group whose relational features follow those of the training set considering the posterior probabilities obtained from support vector machine (SVM), discriminant subspace ensemble (DSE), and artificial neural network (ANN) classifiers. These classifiers are trained on nonrelational features with a 10-fold cross-validation scheme. The anomaly detection module is a bank of ANN autoencoders, each corresponding with an organ, trained on nonrelational features. A heuristic rule detects anomalous organs that exceed predefined organ-specific tolerances for the feature reconstruction error and the classifier's posterior probabilities. Independent data sets with anomalous delineations were used to test the overall performance of the KBQC system. The anomalous delineations were manually manipulated, computer-generated, or propagated based on a transformation obtained by imperfect registrations. Both peer-review-based scoring system and shape similarity coefficient (DSC) were used to label regions of interest (ROIs) as normal or anomalous in two independent test cohorts.

RESULTS

The accuracy of the classifiers was ≥ $\ge$ 99.8%, and the minimum per-class F1-scores were 0.99, 0.99, and 0.98 for SVM, DSE, and ANN, respectively. The group-wise inference system reduced the miss-classification likelihood for the test data set with anomalous delineations compared to each individual classifier and a fused classifier that used the average posterior probability of all classifiers. For 15 independent locally advanced lung patients, the system detected > $>$ 79% of the anomalous ROIs. For 1320 auto-segmented abdominopelvic organs, the anomaly detection system identified anomalous delineations, which also had low Dice similarity coefficient values with respect to manually delineated organs in the training data set.

CONCLUSION

The KBQC system detected anomalous delineations with superior accuracy compared to classification methods that judge only based on posterior probabilities.

Deep learning-based GTV contouring modeling inter- and intra- observer variability in sarcomas

BACKGROUND AND PURPOSE

The delineation of the gross tumor volume (GTV) is a critical step for radiation therapy treatment planning. The delineation procedure is typically performed manually which exposes two major issues: cost and reproducibility. Delineation is a time-consuming process that is subject to inter- and intra-observer variability. While methods have been proposed to predict GTV contours, typical approaches ignore variability and therefore fail to utilize the valuable confidence information offered by multiple contours.

MATERIALS AND METHODS

In this work we propose an automatic GTV contouring method for soft-tissue sarcomas from X-ray computed tomography (CT) images, using deep learning by integrating inter- and intra-observer variability in the learned model. Sixty-eight patients with soft tissue and bone sarcomas were considered in this evaluation, all underwent pre-operative CT imaging used to perform GTV delineation. Four radiation oncologists and radiologists performed three contouring trials each for all patients. We quantify variability by defining confidence levels based on the frequency of inclusion of a given voxel into the GTV and use a deep convolutional neural network to learn GTV confidence maps.

RESULTS

Results were compared to confidence maps from the four readers as well as ground-truth consensus contours established jointly by all readers. The resulting continuous Dice score between predicted and true confidence maps was 87% and the Hausdorff distance was 14 mm.

CONCLUSION

Results demonstrate the ability of the proposed method to predict accurate contours while utilizing variability and as such it can be used to improve clinical workflow.

Evaluating Automatic Segmentation for Swallowing-Related Organs for Head and Neck Cancer

Purpose: To evaluate the accuracy of deep-learning-based auto-segmentation of the superior constrictor, middle constrictor, inferior constrictor, and larynx in comparison with a traditional multi-atlas-based method. Methods and Materials: One hundred and five computed tomography image datasets from 83 head and neck cancer patients were retrospectively collected and the superior constrictor, middle constrictor, inferior constrictor, and larynx were analyzed for deep-learning versus multi-atlas-based segmentation. Eighty-three computed tomography images (40 diagnostic computed tomography and 43 planning computed tomography) were used for training the convolutional neural network, and for atlas-based model training. The remaining 22 computed tomography datasets were used for validation of the atlas-based auto-segmentation versus deep-learning-based auto-segmentation contours, both of which were compared with the corresponding manual contours. Quantitative measures included Dice similarity coefficient, recall, precision, Hausdorff distance, 95th percentile of Hausdorff distance, and mean surface distance. Dosimetric differences between the auto-generated contours and manual contours were evaluated. Subjective evaluation was obtained from 3 clinical observers to blindly score the autosegmented structures based on the percentage of slices that require manual modification. Results: The deep-learning-based auto-segmentation versus atlas-based auto-segmentation results were compared for the superior constrictor, middle constrictor, inferior constrictor, and larynx. The mean Dice similarity coefficient values for the 4 structures were 0.67, 0.60, 0.65, and 0.84 for deep-learning-based auto-segmentation, whereas atlas-based auto-segmentation has Dice similarity coefficient results at 0.45, 0.36, 0.50, and 0.70, respectively. The mean 95th percentile of Hausdorff distance (cm) for the 4 structures were 0.41, 0.57, 0.59, and 0.54 for deep-learning-based auto-segmentation, but 0.78, 0.95, 0.96, and 1.23 for atlas-based auto-segmentation results, respectively. Similar mean dose differences were obtained from the 2 sets of autosegmented contours compared to manual contours. The dose–volume discrepancies and the average modification rates were higher with the atlas-based auto-segmentation contours. Conclusion: Swallowing-related structures are more accurately generated with DL-based versus atlas-based segmentation when compared with manual contours.

Clinical utility of convolutional neural networks for treatment planning in radiotherapy for spinal metastases

•

We presented a CNN workflow for segmentation and labeling of vertebrae on CT.

•

This approach proved to be robust in a majority of cases with spinal metastases.

•

The presented workflow can save time in a clinical radiotherapy setting.

•

The approach also allows for more advanced quantitative image analysis of vertebrae.

Background and purpose

Spine delineation is essential for high quality radiotherapy treatment planning of spinal metastases. However, manual delineation is time-consuming and prone to interobserver variability. Automatic spine delineation, especially using deep learning, has shown promising results in healthy subjects. We aimed to evaluate the clinical utility of deep learning-based vertebral body delineations for radiotherapy planning purposes.

Materials and methods

A multi-scale convolutional neural network (CNN) was used for automatic segmentation and labeling. Two approaches were tested: the combined approach using one CNN for both segmentation and labeling, and the sequential approach using separate CNN’s for these tasks. Training and internal validation data included 580 vertebrae, external validation data included 202 vertebrae. For quantitative assessment, Dice similarity coefficient (DSC) and Hausdorff distance (HD) were used. Axial slices from external images were presented to radiation oncologists for subjective evaluation.

Results

Both approaches performed comparably during the internal validation (DSC: 96.7%, HD: 3.6 mm), but the sequential approach proved more robust during the external validation (DSC: 94.5% vs 94.4%, p < 0.001, HD: 4.5 vs 7.1 mm, p < 0.001). Subsequently, subjective evaluation of this sequential approach showed that experienced radiation oncologists could distinguish automatic from human-made contours in 63% of cases. They rated automatic contours clinically acceptable in 77% of cases, compared to 88% of human-made contours.

Conclusion

We present a feasible approach for automatic vertebral body delineation using two variants of a multi-scale CNN. This approach generates high quality automatic delineations, which can save time in a clinical radiotherapy workflow.

The impact of radiomics in diagnosis and staging of pancreatic cancer

INTRODUCTION

Pancreatic cancer (PC) is one of the most aggressive tumours, and better risk stratification among patients is required to provide tailored treatment. The meaning of radiomics and texture analysis as predictive techniques are not already systematically assessed. The aim of this study is to assess the role of radiomics in PC.

METHODS

A PubMed/MEDLINE and Embase systematic review was conducted to assess the role of radiomics in PC. The search strategy was 'radiomics [All Fields] AND ("pancreas" [MeSH Terms] OR "pancreas" [All Fields] OR "pancreatic" [All Fields])' and only original articles referred to PC in humans in the English language were considered.

RESULTS

A total of 123 studies and 183 studies were obtained using the mentioned search strategy on PubMed and Embase, respectively. After the complete selection process, a total of 56 papers were considered eligible for the analysis of the results. Radiomics methods were applied in PC for assessment technical feasibility and reproducibility aspects analysis, risk stratification, biologic or genomic status prediction and treatment response prediction.

DISCUSSION

Radiomics seems to be a promising approach to evaluate PC from diagnosis to treatment response prediction. Further and larger studies are required to confirm the role and allowed to include radiomics parameter in a comprehensive decision support system.

A comparative study of auto-contouring softwares in delineation of organs at risk in lung cancer and rectal cancer

Radiotherapy requires the target area and the organs at risk to be contoured on the CT image of the patient. During the process of organs-at-Risk (OAR) of the chest and abdomen, the doctor needs to contour at each CT image. The delineations of large and varied shapes are time-consuming and laborious. This study aims to evaluate the results of two automatic contouring softwares on OARs definition of CT images of lung cancer and rectal cancer patients. The CT images of 15 patients with rectal cancer and 15 patients with lung cancer were selected separately, and the organs at risk were manually contoured by experienced physicians as reference structures. And then the same datasets were automatically contoured based on AiContour (version 3.1.8.0, Manufactured by Linking MED, Beijing, China) and Raystation (version 4.7.5.4, Manufactured by Raysearch, Stockholm, Sweden) respectively. Deep learning auto-segmentations and Atlas were respectively performed with AiContour and Raystation. Overlap index (OI), Dice similarity index (DSC) and Volume difference (Dv) were evaluated based on the auto-contours, and independent-sample t-test analysis is applied to the results. The results of deep learning auto-segmentations on OI and DSC were better than that of Atlas with statistical difference. There was no significant difference in Dv between the results of two software. With deep learning auto-segmentations, auto-contouring results of most organs in the chest and abdomen are good, and with slight modification, it can meet the clinical requirements for planning. With Atlas, auto-contouring results in most OAR is not as good as deep learning auto-segmentations, and only the auto-contouring results of some organs can be used clinically after modification.

Metrics to evaluate the performance of auto-segmentation for radiation treatment planning: A critical review

Advances in artificial intelligence-based methods have led to the development and publication of numerous systems for auto-segmentation in radiotherapy. These systems have the potential to decrease contour variability, which has been associated with poor clinical outcomes and increased efficiency in the treatment planning workflow. However, there are no uniform standards for evaluating auto-segmentation platforms to assess their efficacy at meeting these goals. Here, we review the most frequently used evaluation techniques which include geometric overlap, dosimetric parameters, time spent contouring, and clinical rating scales. These data suggest that many of the most commonly used geometric indices, such as the Dice Similarity Coefficient, are not well correlated with clinically meaningful endpoints. As such, a multi-domain evaluation, including composite geometric and/or dosimetric metrics with physician-reported assessment, is necessary to gauge the clinical readiness of auto-segmentation for radiation treatment planning.

Artificial Intelligence in magnetic Resonance guided Radiotherapy: Medical and physical considerations on state of art and future perspectives

Over the last years, technological innovation in Radiotherapy (RT) led to the introduction of Magnetic Resonance-guided RT (MRgRT) systems. Due to the higher soft tissue contrast compared to on-board CT-based systems, MRgRT is expected to significantly improve the treatment in many situations. MRgRT systems may extend the management of inter- and intra-fraction anatomical changes, offering the possibility of online adaptation of the dose distribution according to daily patient anatomy and to directly monitor tumor motion during treatment delivery by means of a continuous cine MR acquisition. Online adaptive treatments require a multidisciplinary and well-trained team, able to perform a series of operations in a safe, precise and fast manner while the patient is waiting on the treatment couch. Artificial Intelligence (AI) is expected to rapidly contribute to MRgRT, primarily by safely and efficiently automatising the various manual operations characterizing online adaptive treatments. Furthermore, AI is finding relevant applications in MRgRT in the fields of image segmentation, synthetic CT reconstruction, automatic (on-line) planning and the development of predictive models based on daily MRI. This review provides a comprehensive overview of the current AI integration in MRgRT from a medical physicist's perspective. Medical physicists are expected to be major actors in solving new tasks and in taking new responsibilities: their traditional role of guardians of the new technology implementation will change with increasing emphasis on the managing of AI tools, processes and advanced systems for imaging and data analysis, gradually replacing many repetitive manual tasks.

Artificial intelligence (AI) and interventional radiotherapy (brachytherapy): state of art and future perspectives

Purpose

Artificial intelligence (AI) plays a central role in building decision supporting systems (DSS), and its application in healthcare is rapidly increasing. The aim of this study was to define the role of AI in healthcare, with main focus on radiation oncology (RO) and interventional radiotherapy (IRT, brachytherapy).

Artificial intelligence in interventional radiation therapy

AI in RO has a large impact in providing clinical decision support, data mining and advanced imaging analysis, automating repetitive tasks, optimizing time, and modelling patients and physicians' behaviors in heterogeneous contexts. Implementing AI and automation in RO and IRT can successfully facilitate all the steps of treatment workflow, such as patient consultation, target volume delineation, treatment planning, and treatment delivery.

Conclusions

AI may contribute to improve clinical outcomes through the application of predictive models and DSS optimization. This approach could lead to reducing time-consuming repetitive tasks, healthcare costs, and improving treatment quality assurance and patient's assistance in IRT.

Assessment of manual adjustment performed in clinical practice following deep learning contouring for head and neck organs at risk in radiotherapy

BACKGROUND AND PURPOSE

Auto-contouring performance has been widely studied in development and commissioning studies in radiotherapy, and its impact on clinical workflow assessed in that context. This study aimed to evaluate the manual adjustment of auto-contouring in routine clinical practice and to identify improvements regarding the auto-contouring model and clinical user interaction, to improve the efficiency of auto-contouring.

MATERIALS AND METHODS

A total of 103 clinical head and neck cancer cases, contoured using a commercial deep-learning contouring system and subsequently checked and edited for clinical use were retrospectively taken from clinical data over a twelve-month period (April 2019-April 2020). The amount of adjustment performed was calculated, and all cases were registered to a common reference frame for assessment purposes. The median, 10th and 90th percentile of adjustment were calculated and displayed using 3D renderings of structures to visually assess systematic and random adjustment. Results were also compared to inter-observer variation reported previously. Assessment was performed for both the whole structures and for regional sub-structures, and according to the radiation therapy technologist (RTT) who edited the contour.

RESULTS

The median amount of adjustment was low for all structures (<2 mm), although large local adjustment was observed for some structures. The median was systematically greater or equal to zero, indicating that the auto-contouring tends to under-segment the desired contour.

CONCLUSION

Auto-contouring performance assessment in routine clinical practice has identified systematic improvements required technically, but also highlighted the need for continued RTT training to ensure adherence to guidelines.

Overview of artificial intelligence-based applications in radiotherapy: Recommendations for implementation and quality assurance

Artificial Intelligence (AI) is currently being introduced into different domains, including medicine. Specifically in radiation oncology, machine learning models allow automation and optimization of the workflow. A lack of knowledge and interpretation of these AI models can hold back wide-spread and full deployment into clinical practice. To facilitate the integration of AI models in the radiotherapy workflow, generally applicable recommendations on implementation and quality assurance (QA) of AI models are presented. For commonly used applications in radiotherapy such as auto-segmentation, automated treatment planning and synthetic computed tomography (sCT) the basic concepts are discussed in depth. Emphasis is put on the commissioning, implementation and case-specific and routine QA of AI models needed for a methodical introduction in clinical practice.

External validation of deep learning-based contouring of head and neck organs at risk

BACKGROUND AND PURPOSE

Head and neck (HN) radiotherapy can benefit from automatic delineation of tumor and surrounding organs because of the complex anatomy and the regular need for adaptation. The aim of this study was to assess the performance of a commercially available deep learning contouring (DLC) model on an external validation set.

MATERIALS AND METHODS

The CT-based DLC model, trained at the University Medical Center Groningen (UMCG), was applied to an independent set of 58 patients from the Radboud University Medical Center (RUMC). DLC results were compared to the RUMC manual reference using the Dice similarity coefficient (DSC) and 95th percentile of Hausdorff distance (HD95). Craniocaudal spatial information was added by calculating binned measures. In addition, a qualitative evaluation compared the acceptance of manual and DLC contours in both groups of observers.

RESULTS

Good correspondence was shown for the mandible (DSC 0.90; HD95 3.6 mm). Performance was reasonable for the glandular OARs, brainstem and oral cavity (DSC 0.78-0.85, HD95 3.7-7.3 mm). The other aerodigestive tract OARs showed only moderate agreement (DSC 0.53-0.65, HD95 around 9 mm). The binned measures displayed the largest deviations caudally and/or cranially.

CONCLUSIONS

This study demonstrates that the DLC model can provide a reasonable starting point for delineation when applied to an independent patient cohort. The qualitative evaluation did not reveal large differences in the interpretation of contouring guidelines between RUMC and UMCG observers.

Implementing user-defined atlas-based auto-segmentation for a large multi-centre organisation: the Australian Experience

INTRODUCTION

Contouring has become an increasingly important aspect of radiation therapy due to inverse planning, and yet is extremely time-consuming. To improve contouring efficiency and reduce potential inter-observer variation, the atlas-based auto-segmentation (ABAS) function in Velocity was introduced to ICON cancer centres (ICC) throughout Australia as a solution for automatic contouring.

METHODS

This paper described the implementation process of the ABAS function and the construction of user-defined atlas sets and compared the contouring efficiency before and after the introduction of ABAS.

RESULTS

The results indicate that the main limitation to the ABAS performance was Velocity's sub-optimal atlas selection method. Three user-defined atlas sets were constructed. Results suggested that the introduction of the ABAS saved at least 5 minutes of manual contouring time (P < 0.05), although further verification was required due to limitations in the data collection method. The pilot rollout adopting a 'champion' approach was successful and provided an opportunity to improve the user-defined atlases prior to the national implementation.

CONCLUSION

The implementation of user-defined ABAS for head and neck (H&N) and female thorax patients at ICCs was successful, which achieved at least 5 minutes of efficiency gain.

Deep Learning: A Review for the Radiation Oncologist

Introduction: Deep Learning (DL) is a machine learning technique that uses deep neural networks to create a model. The application areas of deep learning in radiation oncology include image segmentation and detection, image phenotyping, and radiomic signature discovery, clinical outcome prediction, image dose quantification, dose-response modeling, radiation adaptation, and image generation. In this review, we explain the methods used in DL and perform a literature review using the Medline database to identify studies using deep learning in radiation oncology. The search was conducted in April 2018, and identified studies published between 1997 and 2018, strongly skewed toward 2015 and later. Methods: A literature review was performed using PubMed/Medline in order to identify important recent publications to be synthesized into a review of the current status of Deep Learning in radiation oncology, directed at a clinically-oriented reader. The search strategy included the search terms "radiotherapy" and "deep learning." In addition, reference lists of selected articles were hand searched for further potential hits of relevance to this review. The search was conducted in April 2018, and identified studies published between 1997 and 2018, strongly skewed toward 2015 and later. Results: Studies using DL for image segmentation were identified in Brain (n = 2), Head and Neck (n = 3), Lung (n = 6), Abdominal (n = 2), and Pelvic (n = 6) cancers. Use of Deep Learning has also been reported for outcome prediction, such as toxicity modeling (n = 3), treatment response and survival (n = 2), or treatment planning (n = 5). Conclusion: Over the past few years, there has been a significant number of studies assessing the performance of DL techniques in radiation oncology. They demonstrate how DL-based systems can aid clinicians in their daily work, be it by reducing the time required for or the variability in segmentation, or by helping to predict treatment outcomes and toxicities. It still remains to be seen when these techniques will be employed in routine clinical practice.

Feasibility of multi-atlas cardiac segmentation from thoracic planning CT in a probabilistic framework

Toxicity to cardiac and coronary structures is an important late morbidity for patients undergoing left-sided breast radiotherapy. Many current studies have relied on estimates of cardiac doses assuming standardised anatomy, with a calculated increase in relative risk of 7.4% per Gy (mean heart dose). To provide individualised estimates for dose, delineation of various cardiac structures on patient images is required. Automatic multi-atlas based segmentation can provide a consistent, robust solution, however there are challenges to this method. We are aiming to develop and validate a cardiac atlas and segmentation framework, with a focus on the limitations and uncertainties in the process. We present a probabilistic approach to segmentation, which provides a simple method to incorporate inter-observer variation, as well as a useful tool for evaluating the accuracy and sources of error in segmentation. A dataset consisting of 20 planning computed tomography (CT) images of Australian breast cancer patients with delineations of 17 structures (including whole heart, four chambers, coronary arteries and valves) was manually contoured by three independent observers, following a protocol based on a published reference atlas, with verification by a cardiologist. To develop and validate the segmentation framework a leave-one-out cross-validation strategy was implemented. Performance of the automatic segmentations was evaluated relative to inter-observer variability in manually-derived contours; measures of volume and surface accuracy (Dice similarity coefficient (DSC) and mean absolute surface distance (MASD), respectively) were used to compare automatic segmentation to the consensus segmentation from manual contours. For the whole heart, the resulting segmentation achieved a DSC of [Formula: see text], with a MASD of [Formula: see text] mm. Quantitative results, together with the analysis of probabilistic labelling, indicate the feasibility of accurate and consistent segmentation of larger structures, whereas this is not the case for many smaller structures, where a major limitation in segmentation accuracy is the inter-observer variability in manual contouring.

The impact of training and professional collaboration on the interobserver variation of lung cancer delineations: a multi-institutional study

BACKGROUND

To assess the impact of training and interprofessional collaboration on the interobserver variation in the delineation of the lung gross tumor volume (GTVp) and lymph node (GTVln).

MATERIAL AND METHODS

Eight target volume delineations courses were organized between 2008 and 2013. Specialists and trainees in radiation oncology were asked to delineate the GTVp and GTVln on four representative CT images of a patient diagnosed with lung cancer individually prior each course (baseline), together as group (interprofessional collaboration) and post-training. The mean delineated volume and local standard deviation (local SD) between the contours for each course group were calculated and compared with the expert delineations.

RESULTS

A total 410 delineations were evaluated. The average local SD was lowest for the interprofessional collaboration (GTVp = 0.194 cm, GTVln = 0.371 cm) followed by the post-training (GTVp = 0.244 cm, GTVln = 0.607 cm) and baseline delineations (GTVp = 0.274 cm, GTVln: 0.718 cm). The mean delineated volume was smallest for the interprofessional (GTVp = 4.93 cm³, GTVln = 4.34 cm³) followed by the post-training (GTVp = 5.68 cm³, GTVln = 5.47 cm³) and baseline delineations (GTVp = 6.65 cm³, GTVln = 6.93 cm³). All delineations were larger than the expert for both GTVp and GTVln (p < .001).

CONCLUSION

Our findings indicate that image interpretational differences can lead to large interobserver variation particularly when delineating the GTVln. Interprofessional collaboration was found to have the greatest impact on reducing interobserver variation in the delineation of the GTVln. This highlights the need to develop a clinical workflow so as to ensure that difficult cases are reviewed routinely by a second radiation oncologist or radiologist so as to minimize the risk of geographical tumor miss and unnecessary irradiation to normal tissue.

Assessing the Role of Artificial Intelligence (AI) in Clinical Oncology: Utility of Machine Learning in Radiotherapy Target Volume Delineation

The fields of radiotherapy and clinical oncology have been rapidly changed by the advances of technology. Improvement in computer processing power and imaging quality heralded precision radiotherapy allowing radiotherapy to be delivered efficiently, safely and effectively for patient benefit. Artificial intelligence (AI) is an emerging field of computer science which uses computer models and algorithms to replicate human-like intelligence and perform specific tasks which offers a huge potential to healthcare. We reviewed and presented the history, evolution and advancement in the fields of radiotherapy, clinical oncology and machine learning. Radiotherapy target delineation is a complex task of outlining tumour and organ at risks volumes to allow accurate delivery of radiotherapy. We discussed the radiotherapy planning, treatment delivery and reviewed how technology can help with this challenging process. We explored the evidence and clinical application of machine learning to radiotherapy. We concluded on the challenges, possible future directions and potential collaborations to achieve better outcome for cancer patients.

Comparative evaluation of autocontouring in clinical practice: A practical method using the Turing test

PURPOSE

Automated techniques for estimating the contours of organs and structures in medical images have become more widespread and a variety of measures are available for assessing their quality. Quantitative measures of geometric agreement, for example, overlap with a gold-standard delineation, are popular but may not predict the level of clinical acceptance for the contouring method. Therefore, surrogate measures that relate more directly to the clinical judgment of contours, and to the way they are used in routine workflows, need to be developed. The purpose of this study is to propose a method (inspired by the Turing Test) for providing contour quality measures that directly draw upon practitioners' assessments of manual and automatic contours. This approach assumes that an inability to distinguish automatically produced contours from those of clinical experts would indicate that the contours are of sufficient quality for clinical use. In turn, it is anticipated that such contours would receive less manual editing prior to being accepted for clinical use. In this study, an initial assessment of this approach is performed with radiation oncologists and therapists.

METHODS

Eight clinical observers were presented with thoracic organ-at-risk contours through a web interface and were asked to determine if they were automatically generated or manually delineated. The accuracy of the visual determination was assessed, and the proportion of contours for which the source was misclassified recorded. Contours of six different organs in a clinical workflow were for 20 patient cases. The time required to edit autocontours to a clinically acceptable standard was also measured, as a gold standard of clinical utility. Established quantitative measures of autocontouring performance, such as Dice similarity coefficient with respect to the original clinical contour and the misclassification rate accessed with the proposed framework, were evaluated as surrogates of the editing time measured.

RESULTS

The misclassification rates for each organ were: esophagus 30.0%, heart 22.9%, left lung 51.2%, right lung 58.5%, mediastinum envelope 43.9%, and spinal cord 46.8%. The time savings resulting from editing the autocontours compared to the standard clinical workflow were 12%, 25%, 43%, 77%, 46%, and 50%, respectively, for these organs. The median Dice similarity coefficients between the clinical contours and the autocontours were 0.46, 0.90, 0.98, 0.98, 0.94, and 0.86, respectively, for these organs.

CONCLUSIONS

A better correspondence with time saving was observed for the misclassification rate than the quantitative contour measures explored. From this, we conclude that the inability to accurately judge the source of a contour indicates a reduced need for editing and therefore a greater time saving overall. Hence, task-based assessments of contouring performance may be considered as an additional way of evaluating the clinical utility of autosegmentation methods.

Automatic segmentation software in locally advanced rectal cancer: READY (REsearch program in Auto Delineation sYstem)-RECTAL 02: prospective study

To validate autocontouring software (AS) in a clinical practice including a two steps delineation quality assurance (QA) procedure.

The existing delineation agreement among experts for rectal cancer and the overlap and time criteria that have to be verified to allow the use of AS were defined.

Median Dice Similarity Coefficient (MDSC), Mean slicewise Hausdorff Distances (MSHD) and Total-Time saving (TT) were analyzed.

Two expert Radiation Oncologists reviewed CT-scans of 44 patients and agreed the reference-CTV: the first 14 consecutive cases were used to populate the software Atlas and 30 were used as Test.

Each expert performed a manual (group A) and an automatic delineation (group B) of 15 Test patients.

The delineations were compared with the reference contours.

The overlap between the manual and automatic delineations with MDSC and MSHD and the TT were analyzed.

Three acceptance criteria were set: MDSC ≥ 0.75, MSHD ≤1mm and TT sparing ≥ 50%.

At least 2 criteria had to be met, one of which had to be TT saving, to validate the system.

The MDSC was 0.75, MSHD 2.00 mm and the TT saving 55.5% between group A and group B. MDSC among experts was 0.84.

Autosegmentation systems in rectal cancer partially met acceptability criteria with the present version.

A dose based approach for evaluation of inter-observer variations in target delineation

BACKGROUND AND PURPOSE

Substantial inter-observer variations in target delineation have been presented previously. Target delineation for paediatric cases is difficult due to the small number of children, the variation in paediatric targets, the number of study protocols, and the individual patient's specific needs and demands. Uncertainties in target delineation might lead to under-dosage or over-dosage. The aim of this work is to apply the concept of a consensus volume and good quality treatment plans to visualise and quantify inter-observer target delineation variations in dosimetric terms in addition to conventional geometrically based volume concordance indices.

MATERIAL AND METHODS

Two paediatric cases were used to demonstrate the potential of adding dose metrics when evaluating target delineation diversity; Hodgkin's disease (case 1) and rhabdomyosarcoma of the parotid gland (case 2). The variability in target delineation (PTV delineations) between six centres was quantified using the generalised conformity index, CIgen, generated for volume overlap. The STAPLE algorithm, as implemented in CERR, was used for both cases to derive a consensus volumes. STAPLE is a probabilistic estimate of the true volume generated from all observers. Dose distributions created by each centre for the original target volumes were then applied to this consensus volume.

RESULTS

A considerable variation in target segmentation was seen in both cases. For case 1 the variation was 374-960 cm3 (average 669 cm3) and for case 2; 65-126 cm3 (average 109 cm3). CIgen were 0.53 and 0.70, respectively. The DVHs in absolute volume displayed for the delineated target volume as well as for the consensus volume adds information on both "compliant" target volumes as well as outliers which are hidden with just the use of concordance indices.

CONCLUSIONS

The DVHs in absolute volume add valuable and easily understood information to various indices for evaluating uniformity in target delineation.

Evaluation of rectum and bladder dose accumulation from external beam radiotherapy and brachytherapy for cervical cancer using two different deformable image registration techniques

We evaluated dose-volume histogram (DVH) parameters based on deformable image registration (DIR) between brachytherapy (BT) and external beam radiotherapy (EBRT) that included a center-shielded (CS) plan. Eleven cervical cancer patients were treated with BT, and their pelvic and CS EBRT were studied. Planning CT images for EBRT and BT (except for the first BT, used as the reference image) were deformed with DIR to reference image. We used two DIR parameter settings: intensity-based and hybrid. Mean Dice similarity coefficients (DSCs) comparing EBRT with the reference for the uterus, rectum and bladder were 0.81, 0.77 and 0.83, respectively, for hybrid DIR and 0.47, 0.37 and 0.42, respectively, for intensity-based DIR (P < 0.05). D1 cm3 for hybrid DIR, intensity-based DIR and DVH addition were 75.1, 81.2 and 78.2 Gy, respectively, for the rectum, whereas they were 93.5, 92.3 and 94.3 Gy, respectively, for the bladder. D2 cm3 for hybrid DIR, intensity-based DIR and DVH addition were 70.1, 74.0 and 71.4 Gy, respectively, for the rectum, whereas they were 85.4, 82.8 and 85.4 Gy, respectively, for the bladder. Overall, hybrid DIR obtained higher DSCs than intensity-based DIR, and there were moderate differences in DVH parameters between the two DIR methods, although the results varied among patients. DIR is only experimental, and extra care should be taken when comparing DIR-based dose values with dose-effect curves established using DVH addition. Also, a true evaluation of DIR-based dose accumulation would require ground truth data (e.g. measurement with physical phantom).

Beyond geometrical overlap: a Dosimetrical Evaluation of automated volumes Adaptation (DEA) in head and neck replanning

Introduction

Automated target volumes adaptation could be useful in H&N replanning, but its dosimetric impact has not been analyzed.Primary aim of this investigation is dose coverage assessment in fully automated and edited PTV adaptation settings, compared to manual benchmark.

Materials and methods

Ten IMRT patients were selected and replanning CTs were acquired.A deformable registration with PTV adaptation was performed defining PTVA.PTV B was obtained through manual editing and a benchmark PTV C was manually segmented by a delineation team.The Dice Similarity Index (DSI) and the mean Hausdorff Distance (mHD) were calculated between PTV A and PTV C, and between PTV B and PTV C.One IMRT plan was realized for each PTV: the plans optimized on PTV A and PTV B were proposed on PTV C to evaluate their dosimetric reliability compared to the benchmark plan in terms of PTV V95% dose coverage.

Results

The comparisons between PTV A with PTV C and PTV B with PTV C showed that the better DSI (high) and mHD values (low) are, the smaller difference when compared to PTV C V95% is described.Evaluating plan A and B, PTV C V95% reduced by 6.1 ± 3.0% and by 4.1 ± 2.3% respectively when compared to plan C PTV C V95%.PTV B reaches acceptable dose coverage values (PTV V95% >95%) when DSI is >0.91 and a mHD < 0.17 mm and it has better results when compared to PTV A in 70%.

Discussion

The results show a correlation between the DSI-mHD and the PTV V95% variation, in the comparisons PTV A and PTV B vs PTV C.Furthermore, we observed that PTV V95% coverage is higher in PTV B than in PTV A: the use of automated propagation may not be definitive and requires manual correction.

Dosimetric assessment of an Atlas based automated segmentation for loco-regional radiation therapy of early breast cancer in the Skagen Trial 1: A multi-institutional study

•

40 dose plans from the Skagen Trial 1 collected from Denmark, Belgium and Norway.

•

Atlas-based automated segmentation of each CT scan was obtained using MIM Maestro™.

•

DSC and difference in volume with manual segmentation were collected.

•

HI, V95 and V90% measured on the two different segmentations were compared.

•

Inter-observer variability was low and dose parameters were comparable.

The effect of Atlas-based automated segmentation (ABAS) on dose volume histogram (DVH) parameters compared to manual segmentation (MS) in loco-regional radiotherapy (RT) of early breast cancer was investigated in patients included in the Skagen Trial 1.

This analysis supports implementation of ABAS in clinical practice and multi-institutional trials.

Accuracy of software-assisted contour propagation from planning CT to cone beam CT in head and neck radiotherapy

BACKGROUND

Autocontouring improves workflow in computed tomography (CT)-based dose planning, but could also potentially play a role for optimal use of daily cone beam CT (CBCT) in adaptive radiotherapy. This study aims to determine the accuracy of a deformable image registration (DIR) algorithm for organs at risk (OAR) in the neck region, when applied to CBCT.

MATERIAL AND METHODS

For 30 head and neck cancer (HNC) patients 14 OARs including parotid glands, swallowing structures and spinal cord were delineated. Contours were propagated by DIR from CT to the CBCTs of the first and last treatment fraction. An indirect approach, propagating contours to the first CBCT and from there to the last CBCT was also tested. Propagated contours were compared to manually corrected contours by Dice similarity coefficient (DSC) and Hausdorff distance (HD). Dose was recalculated on CBCTs and dosimetric consequences of uncertainties in DIR were reviewed.

RESULTS

Mean DSC values of ≥0.8 were considered adequate and were achieved in tongue base (0.91), esophagus (0.85), glottic (0.81) and supraglottic larynx (0.83), inferior pharyngeal constrictor muscle (0.84), spinal cord (0.89) and all salivary glands in the first CBCT. For the last CBCT by direct propagation, adequate DSC values were achieved for tongue base (0.85), esophagus (0.84), spinal cord (0.87) and all salivary glands. Using indirect propagation only tongue base (0.80) and parotid glands (0.87) were ≥0.8. Mean relative dose difference between automated and corrected contours was within ±2.5% of planed dose except for esophagus inlet (-4.5%) and esophagus (5.0%) for the last CBCT using indirect propagation.

CONCLUSION

Compared to manually corrected contours, the DIR algorithm was accurate for use in CBCT images of HNC patients and the minor inaccuracies had little consequence for mean dose in most clinically relevant OAR. The method can thus enable a more automated segmentation of CBCT for use in adaptive radiotherapy.

Validation of a Magnetic Resonance Imaging-based Auto-contouring Software Tool for Gross Tumour Delineation in Head and Neck Cancer Radiotherapy Planning

AIMS

To carry out statistical validation of a newly developed magnetic resonance imaging (MRI) auto-contouring software tool for gross tumour volume (GTV) delineation in head and neck tumours to assist in radiotherapy planning.

MATERIALS AND METHODS

Axial MRI baseline scans were obtained for 10 oropharyngeal and laryngeal cancer patients. GTV was present on 102 axial slices and auto-contoured using the modified fuzzy c-means clustering integrated with the level set method (FCLSM). Peer-reviewed (C-gold) manual contours were used as the reference standard to validate auto-contoured GTVs (C-auto) and mean manual contours (C-manual) from two expert clinicians (C1 and C2). Multiple geometric metrics, including the Dice similarity coefficient (DSC), were used for quantitative validation. A DSC≥0.7 was deemed acceptable. Inter- and intra-variabilities among the manual contours were also validated. The two-dimensional contours were then reconstructed in three dimensions for GTV volume calculation, comparison and three-dimensional visualisation.

RESULTS

The mean DSC between C-gold and C-auto was 0.79. The mean DSC between C-gold and C-manual was 0.79 and that between C1 and C2 was 0.80. The average time for GTV auto-contouring per patient was 8 min (range 6-13 min; mean 45 s per axial slice) compared with 15 min (range 6-23 min; mean 88 s per axial slice) for C1. The average volume concordance between C-gold and C-auto volumes was 86.51% compared with 74.16% between C-gold and C-manual. The average volume concordance between C1 and C2 volumes was 86.82%.

CONCLUSIONS

This newly designed MRI-based auto-contouring software tool shows initial acceptable results in GTV delineation of oropharyngeal and laryngeal tumours using FCLSM. This auto-contouring software tool may help reduce inter- and intra-variability and can assist clinical oncologists with time-consuming, complex radiotherapy planning.

Uncertainties in volume delineation in radiation oncology: A systematic review and recommendations for future studies

BACKGROUND AND PURPOSE

Volume delineation is a well-recognised potential source of error in radiotherapy. Whilst it is important to quantify the degree of interobserver variability (IOV) in volume delineation, the resulting impact on dosimetry and clinical outcomes is a more relevant endpoint. We performed a literature review of studies evaluating IOV in target volume and organ-at-risk (OAR) delineation in order to analyse these with respect to the metrics used, reporting of dosimetric consequences, and use of statistical tests.

METHODS AND MATERIALS

Medline and Pubmed databases were queried for relevant articles using keywords. We included studies published in English between 2000 and 2014 with more than two observers.

RESULTS

119 studies were identified covering all major tumour sites. CTV (n=47) and GTV (n=38) were most commonly contoured. Median number of participants and data sets were 7 (3-50) and 9 (1-132) respectively. There was considerable heterogeneity in the use of metrics and methods of analysis. Statistical analysis of results was reported in 68% (n=81) and dosimetric consequences in 21% (n=25) of studies.

CONCLUSION

There is a lack of consistency in conducting and reporting analyses from IOV studies. We suggest a framework to use for future studies evaluating IOV.

Use of auto-segmentation in the delineation of target volumes and organs at risk in head and neck

BACKGROUND

Manual delineation of structures in head and neck cancers is an extremely time-consuming and labor-intensive procedure. With centers worldwide moving towards the use of intensity-modulated radiotherapy and adaptive radiotherapy, there is a need to explore and analyze auto-segmentation (AS) software, in the search for a faster yet accurate method of structure delineation.

MATERIAL AND METHODS

A search for studies published after 2005 comparing AS and manual delineation in contouring organ at risks (OARs) and target volume for head and neck patients was conducted. The reviewed results were then categorized into arguments proposing and opposing the review title.

RESULTS

Ten studies were reviewed and derived results were assessed in terms of delineation time-saving ability and extent of delineation accuracy. The influence of other external factors (observer variability, AS strategies adopted and stage of disease) were also considered. Results were conflicting with some studies demonstrating great potential in replacing manual delineation whereas other studies illustrated otherwise. Six of 10 studies investigated time saving; the largest time saving reported being 59%. However, one study found that additional time of 15.7% was required for AS. Four studies reported AS contours to be between 'reasonably good' and 'better quality' than the clinically used contours. Remaining studies cited lack of contrast, AS strategy used and the need for physician intervention as limitations in the standardized use of AS.

DISCUSSION

The studies demonstrated significant potential of AS as a useful delineation tool in contouring target volumes and OARs in head and neck cancers. However, it is evident that AS cannot totally replace manual delineation in contouring some structures in the head and neck and cannot be used independently without human intervention. It is also emphasized that delineation studies should be conducted locally so as to evaluate the true value of AS in head and neck cancers in a specific center.