An evaluation of MR based deep learning auto-contouring for planning head and neck radiotherapy

HaN-Seg: The head and neck organ-at-risk CT and MR segmentation challenge

BACKGROUND AND PURPOSE

To promote the development of auto-segmentation methods for head and neck (HaN) radiation treatment (RT) planning that exploit the information of computed tomography (CT) and magnetic resonance (MR) imaging modalities, we organized HaN-Seg: The Head and Neck Organ-at-Risk CT and MR Segmentation Challenge.

MATERIALS AND METHODS

The challenge task was to automatically segment 30 organs-at-risk (OARs) of the HaN region in 14 withheld test cases given the availability of 42 publicly available training cases. Each case consisted of one contrast-enhanced CT and one T1-weighted MR image of the HaN region of the same patient, with up to 30 corresponding reference OAR delineation masks. The performance was evaluated in terms of the Dice similarity coefficient (DSC) and 95-percentile Hausdorff distance (HD95), and statistical ranking was applied for each metric by pairwise comparison of the submitted methods using the Wilcoxon signed-rank test.

RESULTS

While 23 teams registered for the challenge, only seven submitted their methods for the final phase. The top-performing team achieved a DSC of 76.9 % and a HD95 of 3.5 mm. All participating teams utilized architectures based on U-Net, with the winning team leveraging rigid MR to CT registration combined with network entry-level concatenation of both modalities.

CONCLUSION

This challenge simulated a real-world clinical scenario by providing non-registered MR and CT images with varying fields-of-view and voxel sizes. Remarkably, the top-performing teams achieved segmentation performance surpassing the inter-observer agreement on the same dataset. These results set a benchmark for future research on this publicly available dataset and on paired multi-modal image segmentation in general.

Assessment of lymph node area coverage with total marrow irradiation and implementation of total marrow and lymphoid irradiation using automated deep learning-based segmentation

BACKGROUND

Total marrow irradiation (TMI) and total marrow and lymphoid irradiation (TMLI) have the advantages. However, delineating target lesions according to TMI and TMLI plans is labor-intensive and time-consuming. In addition, although the delineation of target lesions between TMI and TMLI differs, the clinical distinction is not clear, and the lymph node (LN) area coverage during TMI remains uncertain. Accordingly, this study calculates the LN area coverage according to the TMI plan. Further, a deep learning-based model for delineating LN areas is trained and evaluated.

METHODS

Whole-body regional LN areas were manually contoured in patients treated according to a TMI plan. The dose coverage of the delineated LN areas in the TMI plan was estimated. To train the deep learning model for automatic segmentation, additional whole-body computed tomography data were obtained from other patients. The patients and data were divided into training/validation and test groups and models were developed using the "nnU-NET" framework. The trained models were evaluated using Dice similarity coefficient (DSC), precision, recall, and Hausdorff distance 95 (HD95). The time required to contour and trim predicted results manually using the deep learning model was measured and compared.

RESULTS

The dose coverage for LN areas by TMI plan had V100% (the percentage of volume receiving 100% of the prescribed dose), V95%, and V90% median values of 46.0%, 62.1%, and 73.5%, respectively. The lowest V100% values were identified in the inguinal (14.7%), external iliac (21.8%), and para-aortic (42.8%) LNs. The median values of DSC, precision, recall, and HD95 of the trained model were 0.79, 0.83, 0.76, and 2.63, respectively. The time for manual contouring and simply modified predicted contouring were statistically significantly different.

CONCLUSIONS

The dose coverage in the inguinal, external iliac, and para-aortic LN areas was suboptimal when treatment is administered according to the TMI plan. This research demonstrates that the automatic delineation of LN areas using deep learning can facilitate the implementation of TMLI.

Framework for Radiation Oncology Department-wide Evaluation and Implementation of Commercial Artificial Intelligence Autocontouring

PURPOSE

Artificial intelligence (AI)-based autocontouring in radiation oncology has potential benefits such as standardization and time savings. However, commercial AI solutions require careful evaluation before clinical integration. We developed a multidimensional evaluation method to test pretrained AI-based automated contouring solutions across a network of clinics.

METHODS AND MATERIALS

Curated data included 121 patient planning computed tomography (CT) scans with a total of 859 clinically approved contours used for treatment from 4 clinics. Regions of interest (ROIs) were generated with 3 commercial AI-based automated contouring software solutions (AI1, AI2, AI3) spanning the following disease sites: brain, head and neck (H&N), thorax, abdomen, and pelvis. Quantitative agreement between AI-generated and clinical contours was measured by Dice similarity coefficient (DSC) and Hausdorff distance (HD). Qualitative assessment was performed by multiple experts scoring blinded AI-contours using a Likert scale. Workflow and usability surveying was also conducted.

RESULTS

AI1, AI2, and AI3 contours had high quantitative agreement in 27.8%, 32.8%, and 34.1% of cases (DSC >0.9), performing well in pelvis (median DSC = 0.86/0.88/0.91) and thorax (median DSC = 0.91/0.89/0.91). All 3 solutions had low quantitative agreement in 7.4%, 8.8%, and 6.1% of cases (DSC <0.5), performing worse in brain (median DSC = 0.65/0.78/0.75) and H&N (median DSC = 0.76/0.80/0.81). Qualitatively, AI1 and AI2 contours were acceptable (rated 1-2) with at most minor edits in 70.7% and 74.6% of ROIs (2906 ratings), higher for abdomen (AI1: 79.2%) and thorax (AI2: 90.2%), and lower for H&N (29.0/35.6%). An end-user survey showed strong user preference for full automation and mixed preferences for accuracy versus total number of structures generated.

CONCLUSIONS

Our evaluation method provided a comprehensive analysis of both quantitative and qualitative measures of commercially available pretrained AI autocontouring algorithms. The evaluation framework served as a roadmap for clinical integration that aligned with user workflow preference.

vOARiability: Interobserver and intermodality variability analysis in OAR contouring from head and neck CT and MR images

BACKGROUND

Accurate and consistent contouring of organs-at-risk (OARs) from medical images is a key step of radiotherapy (RT) cancer treatment planning. Most contouring approaches rely on computed tomography (CT) images, but the integration of complementary magnetic resonance (MR) modality is highly recommended, especially from the perspective of OAR contouring, synthetic CT and MR image generation for MR-only RT, and MR-guided RT. Although MR has been recognized as valuable for contouring OARs in the head and neck (HaN) region, the accuracy and consistency of the resulting contours have not been yet objectively evaluated.

PURPOSE

To analyze the interobserver and intermodality variability in contouring OARs in the HaN region, performed by observers with different level of experience from CT and MR images of the same patients.

METHODS

In the final cohort of 27 CT and MR images of the same patients, contours of up to 31 OARs were obtained by a radiation oncology resident (junior observer, JO) and a board-certified radiation oncologist (senior observer, SO). The resulting contours were then evaluated in terms of interobserver variability, characterized as the agreement among different observers (JO and SO) when contouring OARs in a selected modality (CT or MR), and intermodality variability, characterized as the agreement among different modalities (CT and MR) when OARs were contoured by a selected observer (JO or SO), both by the Dice coefficient (DC) and 95-percentile Hausdorff distance (HD95 $_{95}$ ).

RESULTS

The mean (±standard deviation) interobserver variability was 69.0 ± 20.2% and 5.1 ± 4.1 mm, while the mean intermodality variability was 61.6 ± 19.0% and 6.1 ± 4.3 mm in terms of DC and HD95 $_{95}$ , respectively, across all OARs. Statistically significant differences were only found for specific OARs. The performed MR to CT image registration resulted in a mean target registration error of 1.7 ± 0.5 mm, which was considered as valid for the analysis of intermodality variability.

CONCLUSIONS

The contouring variability was, in general, similar for both image modalities, and experience did not considerably affect the contouring performance. However, the results indicate that an OAR is difficult to contour regardless of whether it is contoured in the CT or MR image, and that observer experience may be an important factor for OARs that are deemed difficult to contour. Several of the differences in the resulting variability can be also attributed to adherence to guidelines, especially for OARs with poor visibility or without distinctive boundaries in either CT or MR images. Although considerable contouring differences were observed for specific OARs, it can be concluded that almost all OARs can be contoured with a similar degree of variability in either the CT or MR modality, which works in favor of MR images from the perspective of MR-only and MR-guided RT.

Essentially unedited deep-learning-based OARs are suitable for rigorous oropharyngeal and laryngeal cancer treatment planning

Quality of organ at risk (OAR) autosegmentation is often judged by concordance metrics against the human-generated gold standard. However, the ultimate goal is the ability to use unedited autosegmented OARs in treatment planning, while maintaining the plan quality. We tested this approach with head and neck (HN) OARs generated by a prototype deep-learning (DL) model on patients previously treated for oropharyngeal and laryngeal cancer. Forty patients were selected, with all structures delineated by an experienced physician. For each patient, a set of 13 OARs were generated by the DL model. Each patient was re-planned based on original targets and unedited DL-produced OARs. The new dose distributions were then applied back to the manually delineated structures. The target coverage was evaluated with inhomogeneity index (II) and the relative volume of regret. For the OARs, Dice similarity coefficient (DSC) of areas under the DVH curves, individual DVH objectives, and composite continuous plan quality metric (PQM) were compared. The nearly identical primary target coverage for the original and re-generated plans was achieved, with the same II and relative volume of regret values. The average DSC of the areas under the corresponding pairs of DVH curves was 0.97 ± 0.06. The number of critical DVH points which met the clinical objectives with the dose optimized on autosegmented structures but failed when evaluated on the manual ones was 5 of 896 (0.6%). The average OAR PQM score with the re-planned dose distributions was essentially the same when evaluated either on the autosegmented or manual OARs. Thus, rigorous HN treatment planning is possible with OARs segmented by a prototype DL algorithm with minimal, if any, manual editing.

Evaluation of clinical parallel workflow in online adaptive MR-guided Radiotherapy: A detailed assessment of treatment session times

Introduction

Advancements in MRI-guided radiotherapy (MRgRT) enable clinical parallel workflows (CPW) for online adaptive planning (oART), allowing medical physicists (MPs), physicians (MDs), and radiation therapists (RTTs) to perform their tasks simultaneously. This study evaluates the impact of this upgrade on the total treatment time by analyzing each step of the current 0.35T-MRgRT workflow.

Methods

The time process of the workflow steps for 254 treatment fractions in 0.35 MRgRT was examined. Patients have been grouped based on disease site, breathing modality (BM) (BHI or FB), and fractionation (stereotactic body RT [SBRT] or standard fractionated long course [LC]). The time spent for the following workflow steps in Adaptive Treatment (ADP) was analyzed: Patient Setup Time (PSt), MRI Acquisition and Matching (MRt), MR Re-contouring Time (RCt), Re-Planning Time (RPt), Treatment Delivery Time (TDt). Also analyzed was the timing of treatments that followed a Simple workflow (SMP), without the online re-planning (PSt + MRt + TDt.).

Results

The time analysis revealed that the ADP workflow (median: 34 min) is significantly (p < 0.05) longer than the SMP workflow (19 min). The time required for ADP treatments is significantly influenced by TDt, constituting 40 % of the total time. The oART steps (RCt + RPt) took 11 min (median), representing 27 % of the entire procedure. Overall, 79.2 % of oART fractions were completed in less than 45 min, and 30.6 % were completed in less than 30 min.

Conclusion

This preliminary analysis, along with the comparative assessment against existing literature, underscores the potential of CPW to diminish the overall treatment duration in MRgRT-oART. Additionally, it suggests the potential for CPW to promote a more integrated multidisciplinary approach in the execution of oART.

Artificial intelligence applied to image-guided radiation therapy (IGRT): a systematic review by the Young Group of the Italian Association of Radiotherapy and Clinical Oncology (yAIRO)

INTRODUCTION

The advent of image-guided radiation therapy (IGRT) has recently changed the workflow of radiation treatments by ensuring highly collimated treatments. Artificial intelligence (AI) and radiomics are tools that have shown promising results for diagnosis, treatment optimization and outcome prediction. This review aims to assess the impact of AI and radiomics on modern IGRT modalities in RT.

METHODS

A PubMed/MEDLINE and Embase systematic review was conducted to investigate the impact of radiomics and AI to modern IGRT modalities. The search strategy was "Radiomics" AND "Cone Beam Computed Tomography"; "Radiomics" AND "Magnetic Resonance guided Radiotherapy"; "Radiomics" AND "on board Magnetic Resonance Radiotherapy"; "Artificial Intelligence" AND "Cone Beam Computed Tomography"; "Artificial Intelligence" AND "Magnetic Resonance guided Radiotherapy"; "Artificial Intelligence" AND "on board Magnetic Resonance Radiotherapy" and only original articles up to 01.11.2022 were considered.

RESULTS

A total of 402 studies were obtained using the previously mentioned search strategy on PubMed and Embase. The analysis was performed on a total of 84 papers obtained following the complete selection process. Radiomics application to IGRT was analyzed in 23 papers, while a total 61 papers were focused on the impact of AI on IGRT techniques.

DISCUSSION

AI and radiomics seem to significantly impact IGRT in all the phases of RT workflow, even if the evidence in the literature is based on retrospective data. Further studies are needed to confirm these tools' potential and provide a stronger correlation with clinical outcomes and gold-standard treatment strategies.

Optimising the MR-Linac as a standard treatment modality

The magnetic resonance linear accelerator (MR-Linac) offers a new treatment paradigm, providing improved visualisation of targets and organs at risk while allowing for daily adaptation of treatment plans in real time. Online MR-guided adaptive treatment has reduced treatment uncertainties; however, the additional treatment time and resource requirements may be a concern. We present our experience of integrating an MR-Linac into a busy department and provide recommendations for improved clinical and resource efficiency. Furthermore, we discuss potential future technological innovations that can further optimise clinical productivity in a busy department.

Geometric and dosimetric analysis of CT- and MR-based automatic contouring for the EPTN contouring atlas in neuro-oncology

PURPOSE

Atlas-based and deep-learning contouring (DLC) are methods for automatic segmentation of organs-at-risk (OARs). The European Particle Therapy Network (EPTN) published a consensus-based atlas for delineation of OARs in neuro-oncology. In this study, geometric and dosimetric evaluation of automatically-segmented neuro-oncological OARs was performed using CT- and MR-models following the EPTN-contouring atlas.

METHODS

Image and contouring data from 76 neuro-oncological patients were included. Two atlas-based models (CT-atlas and MR-atlas) and one DLC-model (MR-DLC) were created. Manual contours on registered CT-MR-images were used as ground-truth. Results were analyzed in terms of geometrical (volumetric Dice similarity coefficient (vDSC), surface DSC (sDSC), added path length (APL), and mean slice-wise Hausdorff distance (MSHD)) and dosimetrical accuracy. Distance-to-tumor analysis was performed to analyze to which extent the location of the OAR relative to planning target volume (PTV) has dosimetric impact, using Wilcoxon rank-sum tests.

RESULTS

CT-atlas outperformed MR-atlas for 22/26 OARs. MR-DLC outperformed MR-atlas for all OARs. Highest median (95 %CI) vDSC and sDSC were found for the brainstem in MR-DLC: 0.92 (0.88-0.95) and 0.84 (0.77-0.89) respectively, as well as lowest MSHD: 0.27 (0.22-0.39)cm. Median dose differences (ΔD) were within ± 1 Gy for 24/26(92 %) OARs for all three models. Distance-to-tumor showed a significant correlation for ΔDmax,0.03cc-parameters when splitting the data in ≤ 4 cm and > 4 cm OAR-distance (p < 0.001).

CONCLUSION

MR-based DLC and CT-based atlas-contouring enable high-quality segmentation. It was shown that a combination of both CT- and MR-autocontouring models results in the best quality.

Enhancing Radiotherapy Workflow for Head and Neck Cancer with Artificial Intelligence: A Systematic Review

BACKGROUND

Head and neck cancer (HNC) is characterized by complex-shaped tumors and numerous organs at risk (OARs), inducing challenging radiotherapy (RT) planning, optimization, and delivery. In this review, we provided a thorough description of the applications of artificial intelligence (AI) tools in the HNC RT process.

METHODS

The PubMed database was queried, and a total of 168 articles (2016-2022) were screened by a group of experts in radiation oncology. The group selected 62 articles, which were subdivided into three categories, representing the whole RT workflow: (i) target and OAR contouring, (ii) planning, and (iii) delivery.

RESULTS

The majority of the selected studies focused on the OARs segmentation process. Overall, the performance of AI models was evaluated using standard metrics, while limited research was found on how the introduction of AI could impact clinical outcomes. Additionally, papers usually lacked information about the confidence level associated with the predictions made by the AI models.

CONCLUSIONS

AI represents a promising tool to automate the RT workflow for the complex field of HNC treatment. To ensure that the development of AI technologies in RT is effectively aligned with clinical needs, we suggest conducting future studies within interdisciplinary groups, including clinicians and computer scientists.

3D deep convolution neural network for radiation pneumonitis prediction following stereotactic body radiotherapy

In this study, we investigated 3D convolutional neural networks (CNNs) with input from radiographic and dosimetric datasets of primary lung tumors and surrounding lung volumes to predict the likelihood of radiation pneumonitis (RP). Pre-treatment, 3- and 6-month follow-up computed tomography (CT) and 3D dose datasets from one hundred and ninety-three NSCLC patients treated with stereotactic body radiotherapy (SBRT) were retrospectively collected and analyzed for this study. DenseNet-121 and ResNet-50 models were selected for this study as they are deep neural networks and have been proven to have high accuracy for complex image classification tasks. Both were modified with 3D convolution and max pooling layers to accept 3D datasets. We used a minority class oversampling approach and data augmentation to address the challenges of data imbalance and data scarcity. We built two sets of models for classification of three (No RP, Grade 1 RP, Grade 2 RP) and two (No RP, Yes RP) classes as outputs. The 3D DenseNet-121 models performed better (F1 score [0.81], AUC [0.91] [three class]; F1 score [0.77], AUC [0.84] [two class]) than the 3D ResNet-50 models (F1 score [0.54], AUC [0.72] [three-class]; F1 score [0.68], AUC [0.71] [two-class]) (p = 0.017 for three class predictions). We also attempted to identify salient regions within the input 3D image dataset via integrated gradient (IG) techniques to assess the relevance of the tumor surrounding volume for RP stratification. These techniques appeared to indicate the significance of the tumor and surrounding regions in the prediction of RP. Overall, 3D CNNs performed well to predict clinical RP in our cohort based on the provided image sets and radiotherapy dose information.

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.

Deep learning-based auto segmentation using generative adversarial network on magnetic resonance images obtained for head and neck cancer patients

Purpose

Adaptive radiotherapy requires auto‐segmentation in patients with head and neck (HN) cancer. In the current study, we propose an auto‐segmentation model using a generative adversarial network (GAN) on magnetic resonance (MR) images of HN cancer for MR‐guided radiotherapy (MRgRT).

Material and methods

In the current study, we used a dataset from the American Association of Physicists in Medicine MRI Auto‐Contouring (RT‐MAC) Grand Challenge 2019. Specifically, eight structures in the MR images of HN region, namely submandibular glands, lymph node level II and level III, and parotid glands, were segmented with the deep learning models using a GAN and a fully convolutional network with a U‐net. These images were compared with the clinically used atlas‐based segmentation.

Results

The mean Dice similarity coefficient (DSC) of the U‐net and GAN models was significantly higher than that of the atlas‐based method for all the structures (p < 0.05). Specifically, the maximum Hausdorff distance (HD) was significantly lower than that in the atlas method (p < 0.05). Comparing the 2.5D and 3D U‐nets, the 3D U‐net was superior in segmenting the organs at risk (OAR) for HN patients. The DSC was highest for 0.75–0.85, and the HD was lowest within 5.4 mm of the 2.5D GAN model in all the OARs.

Conclusions

In the current study, we investigated the auto‐segmentation of the OAR for HN patients using U‐net and GAN models on MR images. Our proposed model is potentially valuable for improving the efficiency of HN RT treatment planning.

Deep learning tools for the cancer clinic: an open-source framework with head and neck contour validation

Background

With the rapid growth of deep learning research for medical applications comes the need for clinical personnel to be comfortable and familiar with these techniques. Taking a proven approach, we developed a straightforward open-source framework for producing automatic contours for head and neck planning computed tomography studies using a convolutional neural network (CNN).

Methods

Anonymized studies of 229 patients treated at our clinic for head and neck cancer from 2014 to 2018 were used to train and validate the network. We trained a separate CNN iteration for each of 11 common organs at risk, and then used data from 19 patients previously set aside as test cases for evaluation. We used a commercial atlas-based automatic contouring tool as a comparative benchmark on these test cases to ensure acceptable CNN performance. For the CNN contours and the atlas-based contours, performance was measured using three quantitative metrics and physician reviews using survey and quantifiable correction time for each contour.

Results

The CNN achieved statistically better scores than the atlas-based workflow on the quantitative metrics for 7 of the 11 organs at risk. In the physician review, the CNN contours were more likely to need minor corrections but less likely to need substantial corrections, and the cumulative correction time required was less than for the atlas-based contours for all but two test cases.

Conclusions

With this validation, we packaged the code framework and trained CNN parameters and a no-code, browser-based interface to facilitate reproducibility and expansion of the work. All scripts and files are available in a public GitHub repository and are ready for immediate use under the MIT license. Our work introduces a deep learning tool for automatic contouring that is easy for novice personnel to use.

Supplementary Information

The online version contains supplementary material available at 10.1186/s13014-022-01982-y.

Clinical implementation of deep-learning based auto-contouring tools-Experience of three French radiotherapy centers

Deep-learning (DL)-based auto-contouring solutions have recently been proposed as a convincing alternative to decrease workload of target volumes and organs-at-risk (OAR) delineation in radiotherapy planning and improve inter-observer consistency. However, there is minimal literature of clinical implementations of such algorithms in a clinical routine. In this paper we first present an update of the state-of-the-art of DL-based solutions. We then summarize recent recommendations proposed by the European society for radiotherapy and oncology (ESTRO) to be followed before any clinical implementation of artificial intelligence-based solutions in clinic. The last section describes the methodology carried out by three French radiation oncology departments to deploy CE-marked commercial solutions. Based on the information collected, a majority of OAR are retained by the centers among those proposed by the manufacturers, validating the usefulness of DL-based models to decrease clinicians' workload. Target volumes, with the exception of lymph node areas in breast, head and neck and pelvic regions, whole breast, breast wall, prostate and seminal vesicles, are not available in the three commercial solutions at this time. No implemented workflows are currently available to continuously improve the models, but these can be adapted/retrained in some solutions during the commissioning phase to best fit local practices. In reported experiences, automatic workflows were implemented to limit human interactions and make the workflow more fluid. Recommendations published by the ESTRO group will be of importance for guiding physicists in the clinical implementation of patient specific and regular quality assurances.