Deep learning for segmentation in radiation therapy planning: a review

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.

Deep learning applied to dose prediction in external radiation therapy: A narrative review

Over the last decades, the use of artificial intelligence, machine learning and deep learning in medical fields has skyrocketed. Well known for their results in segmentation, motion management and posttreatment outcome tasks, investigations of machine learning and deep learning models as fast dose calculation or quality assurance tools have been present since 2000. The main motivation for this increasing research and interest in artificial intelligence, machine learning and deep learning is the enhancement of treatment workflows, specifically dosimetry and quality assurance accuracy and time points, which remain important time-consuming aspects of clinical patient management. Since 2014, the evolution of models and architectures for dose calculation has been related to innovations and interest in the theory of information research with pronounced improvements in architecture design. The use of knowledge-based approaches to patient-specific methods has also considerably improved the accuracy of dose predictions. This paper covers the state of all known deep learning architectures and models applied to external radiotherapy with a description of each architecture, followed by a discussion on the performance and future of deep learning predictive models in external radiotherapy.

Deep Learning-Based Techniques in Glioma Brain Tumor Segmentation Using Multi-Parametric MRI: A Review on Clinical Applications and Future Outlooks

This comprehensive review explores the role of deep learning (DL) in glioma segmentation using multiparametric magnetic resonance imaging (MRI) data. The study surveys advanced techniques such as multiparametric MRI for capturing the complex nature of gliomas. It delves into the integration of DL with MRI, focusing on convolutional neural networks (CNNs) and their remarkable capabilities in tumor segmentation. Clinical applications of DL-based segmentation are highlighted, including treatment planning, monitoring treatment response, and distinguishing between tumor progression and pseudo-progression. Furthermore, the review examines the evolution of DL-based segmentation studies, from early CNN models to recent advancements such as attention mechanisms and transformer models. Challenges in data quality, gradient vanishing, and model interpretability are discussed. The review concludes with insights into future research directions, emphasizing the importance of addressing tumor heterogeneity, integrating genomic data, and ensuring responsible deployment of DL-driven healthcare technologies. EVIDENCE LEVEL: N/A TECHNICAL EFFICACY: Stage 2.

Clinical evaluation of a deep learning CBCT auto-segmentation software for prostate adaptive radiation therapy

Purpose

Aim of the present study is to characterize a deep learning-based auto-segmentation software (DL) for prostate cone beam computed tomography (CBCT) images and to evaluate its applicability in clinical adaptive radiation therapy routine.

Materials and methods

Ten patients, who received exclusive radiation therapy with definitive intent on the prostate gland and seminal vesicles, were selected. Femoral heads, bladder, rectum, prostate, and seminal vesicles were retrospectively contoured by four different expert radiation oncologists on patients CBCT, acquired during treatment. Consensus contours (CC) were generated starting from these data and compared with those created by DL with different algorithms, trained on CBCT (DL-CBCT) or computed tomography (DL-CT). Dice similarity coefficient (DSC), centre of mass (COM) shift and volume relative variation (VRV) were chosen as comparison metrics. Since no tolerance limit can be defined, results were also compared with the inter-operator variability (IOV), using the same metrics.

Results

The best agreement between DL and CC was observed for femoral heads (DSC of 0.96 for both DL-CBCT and DL-CT). Performance worsened for low-contrast soft tissue organs: the worst results were found for seminal vesicles (DSC of 0.70 and 0.59 for DL-CBCT and DL-CT, respectively). The analysis shows that it is appropriate to use algorithms trained on the specific imaging modality. Furthermore, the statistical analysis showed that, for almost all considered structures, there is no significant difference between DL-CBCT and human operator in terms of IOV.

Conclusions

The accuracy of DL-CBCT is in accordance with CC; its use in clinical practice is justified by the comparison with the inter-operator variability.

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.

Technical Note: Generalizable and Promptable Artificial Intelligence Model to Augment Clinical Delineation in Radiation Oncology

BACKGROUND

Efficient and accurate delineation of organs at risk (OARs) is a critical procedure for treatment planning and dose evaluation. Deep learning-based auto-segmentation of OARs has shown promising results and is increasingly being used in radiation therapy. However, existing deep learning-based auto-segmentation approaches face two challenges in clinical practice: generalizability and human-AI interaction. A generalizable and promptable auto-segmentation model, which segments OARs of multiple disease sites simultaneously and supports on-the-fly human-AI interaction, can significantly enhance the efficiency of radiation therapy treatment planning.

PURPOSE

Meta's segment anything model (SAM) was proposed as a generalizable and promptable model for next-generation natural image segmentation. We further evaluated the performance of SAM in radiotherapy segmentation.

METHODS

Computed tomography (CT) images of clinical cases from four disease sites at our institute were collected: prostate, lung, gastrointestinal, and head & neck. For each case, we selected the OARs important in radiotherapy treatment planning. We then compared both the Dice coefficients and Jaccard indices derived from three distinct methods: manual delineation (ground truth), automatic segmentation using SAM's 'segment anything' mode, and automatic segmentation using SAM's 'box prompt' mode that implements manual interaction via live prompts during segmentation.

RESULTS

Our results indicate that SAM's segment anything mode can achieve clinically acceptable segmentation results in most OARs with Dice scores higher than 0.7. SAM's box prompt mode further improves Dice scores by 0.1∼0.5. Similar results were observed for Jaccard indices. The results show that SAM performs better for prostate and lung, but worse for gastrointestinal and head & neck. When considering the size of organs and the distinctiveness of their boundaries, SAM shows better performance for large organs with distinct boundaries, such as lung and liver, and worse for smaller organs with less distinct boundaries, like parotid and cochlea.

CONCLUSIONS

Our results demonstrate SAM's robust generalizability with consistent accuracy in automatic segmentation for radiotherapy. Furthermore, the advanced box-prompt method enables the users to augment auto-segmentation interactively and dynamically, leading to patient-specific auto-segmentation in radiation therapy. SAM's generalizability across different disease sites and different modalities makes it feasible to develop a generic auto-segmentation model in radiotherapy.

MS-TCNet: An effective Transformer-CNN combined network using multi-scale feature learning for 3D medical image segmentation

Medical image segmentation is a fundamental research problem in the field of medical image processing. Recently, the Transformer have achieved highly competitive performance in computer vision. Therefore, many methods combining Transformer with convolutional neural networks (CNNs) have emerged for segmenting medical images. However, these methods cannot effectively capture the multi-scale features in medical images, even though texture and contextual information embedded in the multi-scale features are extremely beneficial for segmentation. To alleviate this limitation, we propose a novel Transformer-CNN combined network using multi-scale feature learning for three-dimensional (3D) medical image segmentation, which is called MS-TCNet. The proposed model utilizes a shunted Transformer and CNN to construct an encoder and pyramid decoder, allowing six different scale levels of feature learning. It captures multi-scale features with refinement at each scale level. Additionally, we propose a novel lightweight multi-scale feature fusion (MSFF) module that can fully fuse the different-scale semantic features generated by the pyramid decoder for each segmentation class, resulting in a more accurate segmentation output. We conducted experiments on three widely used 3D medical image segmentation datasets. The experimental results indicated that our method outperformed state-of-the-art medical image segmentation methods, suggesting its effectiveness, robustness, and superiority. Meanwhile, our model has a smaller number of parameters and lower computational complexity than conventional 3D segmentation networks. The results confirmed that the model is capable of effective multi-scale feature learning and that the learned multi-scale features are useful for improving segmentation performance. We open-sourced our code, which can be found at https://github.com/AustinYuAo/MS-TCNet.

Navigating the nuances: comparative analysis and hyperparameter optimisation of neural architectures on contrast-enhanced MRI for liver and liver tumour segmentation

In medical imaging, accurate segmentation is crucial to improving diagnosis, treatment, or both. However, navigating the multitude of available architectures for automatic segmentation can be overwhelming, making it challenging to determine the appropriate type of architecture and tune the most crucial parameters during dataset optimisation. To address this problem, we examined and refined seven distinct architectures for segmenting the liver, as well as liver tumours, with a restricted training collection of 60 3D contrast-enhanced magnetic resonance images (CE-MRI) from the ATLAS dataset. Included in these architectures are convolutional neural networks (CNNs), transformers, and hybrid CNN/transformer architectures. Bayesian search techniques were used for hyperparameter tuning to hasten convergence to the optimal parameter mixes while also minimising the number of trained models. It was unexpected that hybrid models, which typically exhibit superior performance on larger datasets, would exhibit comparable performance to CNNs. The optimisation of parameters contributed to better segmentations, resulting in an average increase of 1.7% and 5.0% in liver and tumour segmentation Dice coefficients, respectively. In conclusion, the findings of this study indicate that hybrid CNN/transformer architectures may serve as a practical substitute for CNNs even in small datasets. This underscores the significance of hyperparameter optimisation.

Closing the loop for AI-ready radiology

BACKGROUND

 In recent years, AI has made significant advancements in medical diagnosis and prognosis. However, the incorporation of AI into clinical practice is still challenging and under-appreciated. We aim to demonstrate a possible vertical integration approach to close the loop for AI-ready radiology.

METHOD

 This study highlights the importance of two-way communication for AI-assisted radiology. As a key part of the methodology, it demonstrates the integration of AI systems into clinical practice with structured reports and AI visualization, giving more insight into the AI system. By integrating cooperative lifelong learning into the AI system, we ensure the long-term effectiveness of the AI system, while keeping the radiologist in the loop.  RESULTS:  We demonstrate the use of lifelong learning for AI systems by incorporating AI visualization and structured reports. We evaluate Memory Aware-Synapses and Rehearsal approach and find that both approaches work in practice. Furthermore, we see the advantage of lifelong learning algorithms that do not require the storing or maintaining of samples from previous datasets.

CONCLUSION

 In conclusion, incorporating AI into the clinical routine of radiology requires a two-way communication approach and seamless integration of the AI system, which we achieve with structured reports and visualization of the insight gained by the model. Closing the loop for radiology leads to successful integration, enabling lifelong learning for the AI system, which is crucial for sustainable long-term performance.

KEY POINTS

  · The integration of AI systems into the clinical routine with structured reports and AI visualization.. · Two-way communication between AI and radiologists is necessary to enable AI that keeps the radiologist in the loop.. · Closing the loop enables lifelong learning, which is crucial for long-term, high-performing AI in radiology..

Estimating tongue deformation during laryngoscopy using a hybrid FEM-multibody model and intraoperative tracking - a cadaver study

Throat tumour margin control remains difficult due to the tight, enclosed space of the oral and throat regions and the tissue deformation resulting from placement of retractors and scopes during surgery. Intraoperative imaging can help with better localization but is hindered by non-image-compatible surgical instruments, cost, and unavailability. We propose a novel method of using instrument tracking and FEM-multibody modelling to simulate soft tissue deformation in the intraoperative setting, without requiring intraoperative imaging, to improve surgical guidance accuracy. We report our first empirical study, based on four trials of a cadaveric head specimen with full neck anatomy, yields a mean TLE of 10.8 ± 5.5 mm, demonstrating methodological feasibility.

Validation of a deep-learning segmentation model for adult and pediatric head and neck radiotherapy in different patient positions

Background and purpose

Autocontouring for radiotherapy has the potential to significantly save time and reduce interobserver variability. We aimed to assess the performance of a commercial autocontouring model for head and neck (H&N) patients in eight orientations relevant to particle therapy with fixed beam lines, focusing on validation and implementation for routine clinical use.

Materials and methods

Autocontouring was performed on sixteen organs at risk (OARs) for 98 adult and pediatric patients with 137 H&N CT scans in eight orientations. A geometric comparison of the autocontours and manual segmentations was performed using the Hausdorff Distance 95th percentile, Dice Similarity Coefficient (DSC) and surface DSC and compared to interobserver variability where available. Additional qualitative scoring and dose-volume-histogram (DVH) parameters analyses were performed for twenty patients in two positions, consisting of scoring on a 0-3 scale based on clinical usability and comparing the mean (Dmean) and near-maximum (D2%) dose, respectively.

Results

For the geometric analysis, the model performance in head-first-supine straight and hyperextended orientations was in the same range as the interobserver variability. HD95, DSC and surface DSC was heterogeneous in other orientations. No significant geometric differences were found between pediatric and adult autocontours. The qualitative scoring yielded a median score of ≥ 2 for 13/16 OARs while 7/32 DVH parameters were significantly different.

Conclusions

For head-first-supine straight and hyperextended scans, we found that 13/16 OAR autocontours were suited for use in daily clinical practice and subsequently implemented. Further development is needed for other patient orientations before implementation.

Multi-organ segmentation of CT via convolutional neural network: impact of training setting and scanner manufacturer

Objective. Automated organ segmentation on CT images can enable the clinical use of advanced quantitative software devices, but model performance sensitivities must be understood before widespread adoption can occur. The goal of this study was to investigate performance differences between Convolutional Neural Networks (CNNs) trained to segment one (single-class) versus multiple (multi-class) organs, and between CNNs trained on scans from a single manufacturer versus multiple manufacturers.Methods. The multi-class CNN was trained on CT images obtained from 455 whole-body PET/CT scans (413 for training, 42 for testing) taken with Siemens, GE, and Phillips PET/CT scanners where 16 organs were segmented. The multi-class CNN was compared to 16 smaller single-class CNNs trained using the same data, but with segmentations of only one organ per model. In addition, CNNs trained on Siemens-only (N = 186) and GE-only (N = 219) scans (manufacturer-specific) were compared with CNNs trained on data from both Siemens and GE scanners (manufacturer-mixed). Segmentation performance was quantified using five performance metrics, including the Dice Similarity Coefficient (DSC).Results. The multi-class CNN performed well compared to previous studies, even in organs usually considered difficult auto-segmentation targets (e.g., pancreas, bowel). Segmentations from the multi-class CNN were significantly superior to those from smaller single-class CNNs in most organs, and the 16 single-class models took, on average, six times longer to segment all 16 organs compared to the single multi-class model. The manufacturer-mixed approach achieved minimally higher performance over the manufacturer-specific approach.Significance. A CNN trained on contours of multiple organs and CT data from multiple manufacturers yielded high-quality segmentations. Such a model is an essential enabler of image processing in a software device that quantifies and analyzes such data to determine a patient's treatment response. To date, this activity of whole organ segmentation has not been adopted due to the intense manual workload and time required.

Extensive clinical testing of Deep Learning Segmentation models for thorax and breast cancer radiotherapy planning

BACKGROUND

The performance of deep learning segmentation (DLS) models for automatic organ extraction from CT images in the thorax and breast regions was investigated. Furthermore, the readiness and feasibility of integrating DLS into clinical practice were addressed by measuring the potential time savings and dosimetric impact.

MATERIAL AND METHODS

Thirty patients referred to radiotherapy for breast cancer were prospectively included. A total of 23 clinically relevant left- and right-sided organs were contoured manually on CT images according to ESTRO guidelines. Next, auto-segmentation was executed, and the geometric agreement between the auto-segmented and manually contoured organs was qualitatively assessed applying a scale in the range [0-not acceptable, 3-no corrections]. A quantitative validation was carried out by calculating Dice coefficients (DSC) and the 95% percentile of Hausdorff distances (HD95). The dosimetric impact of optimizing the treatment plans on the uncorrected DLS contours, was investigated from a dose coverage analysis using DVH values of the manually delineated contours as references.

RESULTS

The qualitative analysis showed that 93% of the DLS generated OAR contours did not need corrections, except for the heart where 67% of the contours needed corrections. The majority of DLS generated CTVs needed corrections, whereas a minority were deemed not acceptable. Still, using the DLS-model for CTV and heart delineation is on average 14 minutes faster. An average DSC=0.91 and H95=9.8 mm were found for the left and right breasts, respectively. Likewise, and average DSC in the range [0.66, 0.76]mm and HD95 in the range [7.04, 12.05]mm were found for the lymph nodes.

CONCLUSION

The validation showed that the DLS generated OAR contours can be used clinically. Corrections were required to most of the DLS generated CTVs, and therefore warrants more attention before possibly implementing the DLS models clinically.

Quantification of tissue property and perfusion uncertainties in hyperthermia treatment planning: Multianalysis using polynomial chaos expansion

INTRODUCTION

Hyperthermia treatment planning (HTP) tools can guide treatment delivery, particularly with locoregional radiative phased array systems. Uncertainties in tissue and perfusion property values presently lead to quantitative inaccuracy of HTP, leading to sub-optimal treatment. Assessment of these uncertainties would allow for better judgement of the reliability of treatment plans and improve their value for treatment guidance. However, systematically investigating the impact of all uncertainties on treatment plans is a complex, high-dimensional problem and too computationally expensive for traditional Monte Carlo approaches. This study aims to systematically quantify the treatment-plan impact of tissue property uncertainties by investigating their individual contribution to, and combined impact on predicted temperature distributions.

METHODS

A novel Polynomial Chaos Expansion (PCE)-based HTP uncertainty quantification was developed and applied for locoregional hyperthermia of modelled tumours in the pancreatic head, prostate, rectum, and cervix. Patient models were based on the Duke and Ella digital human models. Using Plan2Heat, treatment plans were created to optimise tumour temperature (represented by T90) for treatment using the Alba4D system. For all 25-34 modelled tissues, the impact of tissue property uncertainties was analysed individually i.e., electrical and thermal conductivity, permittivity, density, specific heat capacity and perfusion. Next, combined analyses were performed on the top 30 uncertainties with the largest impact.

RESULTS

Uncertainties in thermal conductivity and heat capacity were found to have negligible impact on the predicted temperature ( < 1 × 10-10 °C), density and permittivity uncertainties had a small impact (< 0.3 °C). Uncertainties in electrical conductivity and perfusion can lead to large variations in predicted temperature. However, variations in muscle properties result in the largest impact at locations that could limit treatment quality, with a standard deviation up to almost 6 °C (pancreas) and 3.5 °C (prostate) for perfusion and electrical conductivity, respectively. The combined influence of all significant uncertainties leads to large variations with a standard deviation up to 9.0, 3.6, 3.7 and 4.1 °C for the pancreatic, prostate, rectal and cervical cases, respectively.

CONCLUSION

Uncertainties in tissue and perfusion property values can have a large impact on predicted temperatures from hyperthermia treatment planning. PCE-based analysis helps to identify all major uncertainties, their impact and judge the reliability of treatment plans.

Seasonal pigment fluctuation in diploid and polyploid Arabidopsis revealed by machine learning-based phenotyping method PlantServation

Long-term field monitoring of leaf pigment content is informative for understanding plant responses to environments distinct from regulated chambers but is impractical by conventional destructive measurements. We developed PlantServation, a method incorporating robust image-acquisition hardware and deep learning-based software that extracts leaf color by detecting plant individuals automatically. As a case study, we applied PlantServation to examine environmental and genotypic effects on the pigment anthocyanin content estimated from leaf color. We processed >4 million images of small individuals of four Arabidopsis species in the field, where the plant shape, color, and background vary over months. Past radiation, coldness, and precipitation significantly affected the anthocyanin content. The synthetic allopolyploid A. kamchatica recapitulated the fluctuations of natural polyploids by integrating diploid responses. The data support a long-standing hypothesis stating that allopolyploids can inherit and combine the traits of progenitors. PlantServation facilitates the study of plant responses to complex environments termed "in natura".

Deep Learning and Registration-Based Mapping for Analyzing the Distribution of Nodal Metastases in Head and Neck Cancer Cohorts: Informing Optimal Radiotherapy Target Volume Design

We introduce a deep-learning- and a registration-based method for automatically analyzing the spatial distribution of nodal metastases (LNs) in head and neck (H/N) cancer cohorts to inform radiotherapy (RT) target volume design. The two methods are evaluated in a cohort of 193 H/N patients/planning CTs with a total of 449 LNs. In the deep learning method, a previously developed nnU-Net 3D/2D ensemble model is used to autosegment 20 H/N levels, with each LN subsequently being algorithmically assigned to the closest-level autosegmentation. In the nonrigid-registration-based mapping method, LNs are mapped into a calculated template CT representing the cohort-average patient anatomy, and kernel density estimation is employed to estimate the underlying average 3D-LN probability distribution allowing for analysis and visualization without prespecified level definitions. Multireader assessment by three radio-oncologists with majority voting was used to evaluate the deep learning method and obtain the ground-truth distribution. For the mapping technique, the proportion of LNs predicted by the 3D probability distribution for each level was calculated and compared to the deep learning and ground-truth distributions. As determined by a multireader review with majority voting, the deep learning method correctly categorized all 449 LNs to their respective levels. Level 2 showed the highest LN involvement (59.0%). The level involvement predicted by the mapping technique was consistent with the ground-truth distribution (p for difference 0.915). Application of the proposed methods to multicenter cohorts with selected H/N tumor subtypes for informing optimal RT target volume design is promising.

Deep learning auto-segmentation on multi-sequence magnetic resonance images for upper abdominal organs

Introduction

Multi-sequence multi-parameter MRIs are often used to define targets and/or organs at risk (OAR) in radiation therapy (RT) planning. Deep learning has so far focused on developing auto-segmentation models based on a single MRI sequence. The purpose of this work is to develop a multi-sequence deep learning based auto-segmentation (mS-DLAS) based on multi-sequence abdominal MRIs.

Materials and methods

Using a previously developed 3DResUnet network, a mS-DLAS model using 4 T1 and T2 weighted MRI acquired during routine RT simulation for 71 cases with abdominal tumors was trained and tested. Strategies including data pre-processing, Z-normalization approach, and data augmentation were employed. Additional 2 sequence specific T1 weighted (T1-M) and T2 weighted (T2-M) models were trained to evaluate performance of sequence-specific DLAS. Performance of all models was quantitatively evaluated using 6 surface and volumetric accuracy metrics.

Results

The developed DLAS models were able to generate reasonable contours of 12 upper abdomen organs within 21 seconds for each testing case. The 3D average values of dice similarity coefficient (DSC), mean distance to agreement (MDA mm), 95 percentile Hausdorff distance (HD95% mm), percent volume difference (PVD), surface DSC (sDSC), and relative added path length (rAPL mm/cc) over all organs were 0.87, 1.79, 7.43, -8.95, 0.82, and 12.25, respectively, for mS-DLAS model. Collectively, 71% of the auto-segmented contours by the three models had relatively high quality. Additionally, the obtained mS-DLAS successfully segmented 9 out of 16 MRI sequences that were not used in the model training.

Conclusion

We have developed an MRI-based mS-DLAS model for auto-segmenting of upper abdominal organs on MRI. Multi-sequence segmentation is desirable in routine clinical practice of RT for accurate organ and target delineation, particularly for abdominal tumors. Our work will act as a stepping stone for acquiring fast and accurate segmentation on multi-contrast MRI and make way for MR only guided radiation therapy.

A convolutional neural network with self-attention for fully automated metabolic tumor volume delineation of head and neck cancer in [Formula: see text]F]FDG PET/CT

PURPOSE

PET-derived metabolic tumor volume (MTV) and total lesion glycolysis of the primary tumor are known to be prognostic of clinical outcome in head and neck cancer (HNC). Including evaluation of lymph node metastases can further increase the prognostic value of PET but accurate manual delineation and classification of all lesions is time-consuming and prone to interobserver variability. Our goal, therefore, was development and evaluation of an automated tool for MTV delineation/classification of primary tumor and lymph node metastases in PET/CT investigations of HNC patients.

METHODS

Automated lesion delineation was performed with a residual 3D U-Net convolutional neural network (CNN) incorporating a multi-head self-attention block. 698 [Formula: see text]F]FDG PET/CT scans from 3 different sites and 5 public databases were used for network training and testing. An external dataset of 181 [Formula: see text]F]FDG PET/CT scans from 2 additional sites was employed to assess the generalizability of the network. In these data, primary tumor and lymph node (LN) metastases were interactively delineated and labeled by two experienced physicians. Performance of the trained network models was assessed by 5-fold cross-validation in the main dataset and by pooling results from the 5 developed models in the external dataset. The Dice similarity coefficient (DSC) for individual delineation tasks and the primary tumor/metastasis classification accuracy were used as evaluation metrics. Additionally, a survival analysis using univariate Cox regression was performed comparing achieved group separation for manual and automated delineation, respectively.

RESULTS

In the cross-validation experiment, delineation of all malignant lesions with the trained U-Net models achieves DSC of 0.885, 0.805, and 0.870 for primary tumor, LN metastases, and the union of both, respectively. In external testing, the DSC reaches 0.850, 0.724, and 0.823 for primary tumor, LN metastases, and the union of both, respectively. The voxel classification accuracy was 98.0% and 97.9% in cross-validation and external data, respectively. Univariate Cox analysis in the cross-validation and the external testing reveals that manually and automatically derived total MTVs are both highly prognostic with respect to overall survival, yielding essentially identical hazard ratios (HR) ([Formula: see text]; [Formula: see text] vs. [Formula: see text]; [Formula: see text] in cross-validation and [Formula: see text]; [Formula: see text] vs. [Formula: see text]; [Formula: see text] in external testing).

CONCLUSION

To the best of our knowledge, this work presents the first CNN model for successful MTV delineation and lesion classification in HNC. In the vast majority of patients, the network performs satisfactory delineation and classification of primary tumor and lymph node metastases and only rarely requires more than minimal manual correction. It is thus able to massively facilitate study data evaluation in large patient groups and also does have clear potential for supervised clinical application.

Comparison of the output of a deep learning segmentation model for locoregional breast cancer radiotherapy trained on 2 different datasets

Introduction

The development of deep learning (DL) models for auto-segmentation is increasing and more models become commercially available. Mostly, commercial models are trained on external data. To study the effect of using a model trained on external data, compared to the same model trained on in-house collected data, the performance of these two DL models was evaluated.

Methods

The evaluation was performed using in-house collected data of 30 breast cancer patients. Quantitative analysis was performed using Dice similarity coefficient (DSC), surface DSC (sDSC) and 95th percentile of Hausdorff Distance (95% HD). These values were compared with previously reported inter-observer variations (IOV).

Results

For a number of structures, statistically significant differences were found between the two models. For organs at risk, mean values for DSC ranged from 0.63 to 0.98 and 0.71 to 0.96 for the in-house and external model, respectively. For target volumes, mean DSC values of 0.57 to 0.94 and 0.33 to 0.92 were found. The difference of 95% HD values ranged 0.08 to 3.23 mm between the two models, except for CTVn4 with 9.95 mm. For the external model, both DSC and 95% HD are outside the range of IOV for CTVn4, whereas this is the case for the DSC found for the thyroid of the in-house model.

Conclusions

Statistically significant differences were found between both models, which were mostly within published inter-observer variations, showing clinical usefulness of both models. Our findings could encourage discussion and revision of existing guidelines, to further decrease inter-observer, but also inter-institute variability.

Simultaneous Integrated Boost (SIB) vs. Sequential Boost in Head and Neck Cancer (HNC) Radiotherapy: A Radiomics-Based Decision Proof of Concept

Artificial intelligence (AI) and in particular radiomics has opened new horizons by extracting data from medical imaging that could be used not only to improve diagnostic accuracy, but also to be included in predictive models contributing to treatment stratification of cancer. Head and neck cancers (HNC) are associated with higher recurrence rates, especially in advanced stages of disease. It is considered that approximately 50% of cases will evolve with loco-regional recurrence, even if they will benefit from a current standard treatment consisting of definitive chemo-radiotherapy. Radiotherapy, the cornerstone treatment in locally advanced HNC, could be delivered either by the simultaneous integrated boost (SIB) technique or by the sequential boost technique, the decision often being a subjective one. The principles of radiobiology could be the basis of an optimal decision between the two methods of radiation dose delivery, but the heterogeneity of HNC radio-sensitivity makes this approach difficult. Radiomics has demonstrated the ability to non-invasively predict radio-sensitivity and the risk of relapse in HNC. Tumor heterogeneity evaluated with radiomics, the inclusion of coarseness, entropy and other first order features extracted from gross tumor volume (GTV) in multivariate models could identify pre-treatment cases that will benefit from one of the approaches (SIB or sequential boost radio-chemotherapy) considered the current standard of care for locally advanced HNC. Computer tomography (CT) simulation and daily cone beam CT (CBCT) could be chosen as imaging source for radiomic analysis.

Systematic Review of Tumor Segmentation Strategies for Bone Metastases

Purpose: To investigate the segmentation approaches for bone metastases in differentiating benign from malignant bone lesions and characterizing malignant bone lesions. Method: The literature search was conducted in Scopus, PubMed, IEEE and MedLine, and Web of Science electronic databases following the guidelines of Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA). A total of 77 original articles, 24 review articles, and 1 comparison paper published between January 2010 and March 2022 were included in the review. Results: The results showed that most studies used neural network-based approaches (58.44%) and CT-based imaging (50.65%) out of 77 original articles. However, the review highlights the lack of a gold standard for tumor boundaries and the need for manual correction of the segmentation output, which largely explains the absence of clinical translation studies. Moreover, only 19 studies (24.67%) specifically mentioned the feasibility of their proposed methods for use in clinical practice. Conclusion: Development of tumor segmentation techniques that combine anatomical information and metabolic activities is encouraging despite not having an optimal tumor segmentation method for all applications or can compensate for all the difficulties built into data limitations.

Deep learning for automatic head and neck lymph node level delineation provides expert-level accuracy

Background

Deep learning-based head and neck lymph node level (HN_LNL) autodelineation is of high relevance to radiotherapy research and clinical treatment planning but still underinvestigated in academic literature. In particular, there is no publicly available open-source solution for large-scale autosegmentation of HN_LNL in the research setting.

Methods

An expert-delineated cohort of 35 planning CTs was used for training of an nnU-net 3D-fullres/2D-ensemble model for autosegmentation of 20 different HN_LNL. A second cohort acquired at the same institution later in time served as the test set (n = 20). In a completely blinded evaluation, 3 clinical experts rated the quality of deep learning autosegmentations in a head-to-head comparison with expert-created contours. For a subgroup of 10 cases, intraobserver variability was compared to the average deep learning autosegmentation accuracy on the original and recontoured set of expert segmentations. A postprocessing step to adjust craniocaudal boundaries of level autosegmentations to the CT slice plane was introduced and the effect of autocontour consistency with CT slice plane orientation on geometric accuracy and expert rating was investigated.

Results

Blinded expert ratings for deep learning segmentations and expert-created contours were not significantly different. Deep learning segmentations with slice plane adjustment were rated numerically higher (mean, 81.0 vs. 79.6, p = 0.185) and deep learning segmentations without slice plane adjustment were rated numerically lower (77.2 vs. 79.6, p = 0.167) than manually drawn contours. In a head-to-head comparison, deep learning segmentations with CT slice plane adjustment were rated significantly better than deep learning contours without slice plane adjustment (81.0 vs. 77.2, p = 0.004). Geometric accuracy of deep learning segmentations was not different from intraobserver variability (mean Dice per level, 0.76 vs. 0.77, p = 0.307). Clinical significance of contour consistency with CT slice plane orientation was not represented by geometric accuracy metrics (volumetric Dice, 0.78 vs. 0.78, p = 0.703).

Conclusions

We show that a nnU-net 3D-fullres/2D-ensemble model can be used for highly accurate autodelineation of HN_LNL using only a limited training dataset that is ideally suited for large-scale standardized autodelineation of HN_LNL in the research setting. Geometric accuracy metrics are only an imperfect surrogate for blinded expert rating.

The role of artificial intelligence based on PET/CT radiomics in NSCLC: Disease management, opportunities, and challenges

Objectives

Lung cancer has been widely characterized through radiomics and artificial intelligence (AI). This review aims to summarize the published studies of AI based on positron emission tomography/computed tomography (PET/CT) radiomics in non-small-cell lung cancer (NSCLC).

Materials and methods

A comprehensive search of literature published between 2012 and 2022 was conducted on the PubMed database. There were no language or publication status restrictions on the search. About 127 articles in the search results were screened and gradually excluded according to the exclusion criteria. Finally, this review included 39 articles for analysis.

Results

Classification is conducted according to purposes and several studies were identified at each stage of disease:1) Cancer detection (n=8), 2) histology and stage of cancer (n=11), 3) metastases (n=6), 4) genotype (n=6), 5) treatment outcome and survival (n=8). There is a wide range of heterogeneity among studies due to differences in patient sources, evaluation criteria and workflow of radiomics. On the whole, most models show diagnostic performance comparable to or even better than experts, and the common problems are repeatability and clinical transformability.

Conclusion

AI-based PET/CT Radiomics play potential roles in NSCLC clinical management. However, there is still a long way to go before being translated into clinical application. Large-scale, multi-center, prospective research is the direction of future efforts, while we need to face the risk of repeatability of radiomics features and the limitation of access to large databases.

Towards interactive deep-learning for tumour segmentation in head and neck cancer radiotherapy

Background and purpose

With deep-learning, gross tumour volume (GTV) auto-segmentation has substantially been improved, but still substantial manual corrections are needed. With interactive deep-learning (iDL), manual corrections can be used to update a deep-learning tool while delineating, minimising the input to achieve acceptable segmentations. We present an iDL tool for GTV segmentation that took annotated slices as input and simulated its performance on a head and neck cancer (HNC) dataset.

Materials and methods

Multimodal image data of 204 HNC patients with clinical tumour and lymph node GTV delineations were used. A baseline convolutional neural network (CNN) was trained (n = 107 training, n = 22 validation) and tested (n = 24). Subsequently, user input was simulated on initial test set by replacing one or more of predicted slices with ground truth delineation, followed by re-training the CNN. The objective was to optimise re-training parameters and simulate slice selection scenarios while limiting annotations to maximally-five slices. The remaining 51 patients were used as an independent test set, where Dice similarity coefficient (DSC), mean surface distance (MSD), and 95% Hausdorff distance (HD_95%) were assessed at baseline and after every update.

Results

Median segmentation accuracy at baseline was DSC = 0.65, MSD = 4.3 mm, HD_95% = 17.5 mm. Updating CNN using three slices equally sampled from the craniocaudal axis of the GTV in the first round, followed by two rounds of annotating one extra slice, gave the best results. The accuracy improved to DSC = 0.82, MSD = 1.6 mm, HD_95% = 4.8 mm. Every CNN update took 30 s.

Conclusions

The presented iDL tool achieved substantial segmentation improvement with only five annotated slices.

Auto-detection and segmentation of involved lymph nodes in HPV-associated oropharyngeal cancer using a convolutional deep learning neural network

•

Patients diagnosed with HPV-associated OPC are often found to have clinical evidence of lymph node involvement at the time of diagnosis.

•

Manual segmentation of radiographically involved lymph nodes is an integral part of treatment planning for those patients dispositioned to definitive radiotherapy.

•

We present a deep learning, convolutional neural network that can be used to automate the process of lymph node detection and segmentation and with potential to increase the efficiency to the radiotherapy workflow.

Purpose

Segmentation of involved lymph nodes on head and neck computed tomography (HN-CT) scans is necessary for the radiotherapy planning of early-stage human papilloma virus (HPV) associated oropharynx cancers (OPC). We aimed to train a deep learning convolutional neural network (DL-CNN) to segment involved lymph nodes on HN-CT scans.

Methods

Ground-truth segmentation of involved nodes was performed on pre-surgical HN-CT scans for 90 patients who underwent levels II-IV neck dissection for node-positive HPV-OPC (training/validation [n = 70] and testing [n = 20]). A 5-fold cross validation approach was used to train 5 DL-CNN sub-models based on a residual U-net architecture. Validation and testing segmentation masks were compared to ground-truth masks using predetermined metrics. A lymph auto-detection model to discriminate between “node-positive” and “node-negative” HN-CT scans was developed by thresholding segmentation model outputs and evaluated using the area under the receiver operating characteristic curve (AUC).

Results

In the DL-CNN validation phase, all sub-models yielded segmentation masks with median Dice ≥ 0.90 and median volume similarity score of ≥ 0.95. In the testing phase, the DL-CNN produced consensus segmentation masks with median Dice of 0.92 (IQR, 0.89–0.95), median volume similarity of 0.97 (IQR, 0.94–0.99), and median Hausdorff distance of 4.52 mm (IQR, 1.22–8.38). The detection model achieved an AUC of 0.98.

Conclusion

The results from this single-institution study demonstrate the successful automation of lymph node segmentation for patients with HPV-OPC using a DL-CNN. Future studies, including validation with an external dataset, are necessary to clarify its role in the larger radiation oncology treatment planning workflow.

Infrastructure platform for privacy-preserving distributed machine learning development of computer-assisted theragnostics in cancer

INTRODUCTION

Emerging evidence suggests that data-driven support tools have found their way into clinical decision-making in a number of areas, including cancer care. Improving them and widening their scope of availability in various differing clinical scenarios, including for prognostic models derived from retrospective data, requires co-ordinated data sharing between clinical centres, secondary analyses of large multi-institutional clinical trial data, or distributed (federated) learning infrastructures. A systematic approach to utilizing routinely collected data across cancer care clinics remains a significant challenge due to privacy, administrative and political barriers.

METHODS

An information technology infrastructure and web service software was developed and implemented which uses machine learning to construct clinical decision support systems in a privacy-preserving manner across datasets geographically distributed in different hospitals. The infrastructure was deployed in a network of Australian hospitals. A harmonized, international ontology-linked, set of lung cancer databases were built with the routine clinical and imaging data at each centre. The infrastructure was demonstrated with the development of logistic regression models to predict major cardiovascular events following radiation therapy.

RESULTS

The infrastructure implemented forms the basis of the Australian computer-assisted theragnostics (AusCAT) network for radiation oncology data extraction, reporting and distributed learning. Four radiation oncology departments (across seven hospitals) in New South Wales (NSW) participated in this demonstration study. Infrastructure was deployed at each centre and used to develop a model predicting for cardiovascular admission within a year of receiving curative radiotherapy for non-small cell lung cancer. A total of 10417 lung cancer patients were identified with 802 being eligible for the model. Twenty features were chosen for analysis from the clinical record and linked registries. After selection, 8 features were included and a logistic regression model achieved an area under the receiver operating characteristic (AUROC) curve of 0.70 and C-index of 0.65 on out-of-sample data.

CONCLUSION

The infrastructure developed was demonstrated to be usable in practice between clinical centres to harmonize routinely collected oncology data and develop models with federated learning. It provides a promising approach to enable further research studies in radiation oncology using real world clinical data.

Scalable radiotherapy data curation infrastructure for deep-learning based autosegmentation of organs-at-risk: A case study in head and neck cancer

In this era of patient-centered, outcomes-driven and adaptive radiotherapy, deep learning is now being successfully applied to tackle imaging-related workflow bottlenecks such as autosegmentation and dose planning. These applications typically require supervised learning approaches enabled by relatively large, curated radiotherapy datasets which are highly reflective of the contemporary standard of care. However, little has been previously published describing technical infrastructure, recommendations, methods or standards for radiotherapy dataset curation in a holistic fashion. Our radiation oncology department has recently embarked on a large-scale project in partnership with an external partner to develop deep-learning-based tools to assist with our radiotherapy workflow, beginning with autosegmentation of organs-at-risk. This project will require thousands of carefully curated radiotherapy datasets comprising all body sites we routinely treat with radiotherapy. Given such a large project scope, we have approached the need for dataset curation rigorously, with an aim towards building infrastructure that is compatible with efficiency, automation and scalability. Focusing on our first use-case pertaining to head and neck cancer, we describe our developed infrastructure and novel methods applied to radiotherapy dataset curation, inclusive of personnel and workflow organization, dataset selection, expert organ-at-risk segmentation, quality assurance, patient de-identification, data archival and transfer. Over the course of approximately 13 months, our expert multidisciplinary team generated 490 curated head and neck radiotherapy datasets. This task required approximately 6000 human-expert hours in total (not including planning and infrastructure development time). This infrastructure continues to evolve and will support ongoing and future project efforts.

[What do we need to deliver "online" adapted radiotherapy treatment plans?]

During the joint SFRO/SFPM session of the 2019 congress, a state of the art of adaptive radiotherapy announced a strong impact in our clinical practice, in particular with the availability of treatment devices coupled to an MRI system. Three years later, it seems relevant to take stock of adaptive radiotherapy in practice, and especially the "online" strategy because it is indeed more and more accessible with recent hardware and software developments, such as coupled accelerators to a three-dimensional imaging device and algorithms based on artificial intelligence. However, the deployment of this promising strategy is complex because it contracts the usual time scale and upsets the usual organizations. So what do we need to deliver adapted treatment plans with an "online" strategy?

Performance of Machine Learning for Tissue Outcome Prediction in Acute Ischemic Stroke: A Systematic Review and Meta-Analysis

Machine learning (ML) has been proposed for lesion segmentation in acute ischemic stroke (AIS). This study aimed to provide a systematic review and meta-analysis of the overall performance of current ML algorithms for final infarct prediction from baseline imaging. We made a comprehensive literature search on eligible studies developing ML models for core infarcted tissue estimation on admission CT or MRI in AIS patients. Eleven studies meeting the inclusion criteria were included in the quantitative analysis. Study characteristics, model methodology, and predictive performance of the included studies were extracted. A meta-analysis was conducted on the dice similarity coefficient (DSC) score by using a random-effects model to assess the overall predictive performance. Study heterogeneity was assessed by Cochrane Q and Higgins I² tests. The pooled DSC score of the included ML models was 0.50 (95% CI 0.39–0.61), with high heterogeneity observed across studies (I² 96.5%, p < 0.001). Sensitivity analyses using the one-study removed method showed the adjusted overall DSC score ranged from 0.47 to 0.52. Subgroup analyses indicated that the DL-based models outperformed the conventional ML classifiers with the best performance observed in DL algorithms combined with CT data. Despite the presence of heterogeneity, current ML-based approaches for final infarct prediction showed moderate but promising performance. Before well integrated into clinical stroke workflow, future investigations are suggested to train ML models on large-scale, multi-vendor data, validate on external cohorts and adopt formalized reporting standards for improving model accuracy and robustness.

Pelvic U-Net: multi-label semantic segmentation of pelvic organs at risk for radiation therapy anal cancer patients using a deeply supervised shuffle attention convolutional neural network

BACKGROUND

Delineation of organs at risk (OAR) for anal cancer radiation therapy treatment planning is a manual and time-consuming process. Deep learning-based methods can accelerate and partially automate this task. The aim of this study was to develop and evaluate a deep learning model for automated and improved segmentations of OAR in the pelvic region.

METHODS

A 3D, deeply supervised U-Net architecture with shuffle attention, referred to as Pelvic U-Net, was trained on 143 computed tomography (CT) volumes, to segment OAR in the pelvic region, such as total bone marrow, rectum, bladder, and bowel structures. Model predictions were evaluated on an independent test dataset (n = 15) using the Dice similarity coefficient (DSC), the 95th percentile of the Hausdorff distance (HD₉₅), and the mean surface distance (MSD). In addition, three experienced radiation oncologists rated model predictions on a scale between 1-4 (excellent, good, acceptable, not acceptable). Model performance was also evaluated with respect to segmentation time, by comparing complete manual delineation time against model prediction time without and with manual correction of the predictions. Furthermore, dosimetric implications to treatment plans were evaluated using different dose-volume histogram (DVH) indices.

RESULTS

Without any manual corrections, mean DSC values of 97%, 87% and 94% were found for total bone marrow, rectum, and bladder. Mean DSC values for bowel cavity, all bowel, small bowel, and large bowel were 95%, 91%, 87% and 81%, respectively. Total bone marrow, bladder, and bowel cavity segmentations derived from our model were rated excellent (89%, 93%, 42%), good (9%, 5%, 42%), or acceptable (2%, 2%, 16%) on average. For almost all the evaluated DVH indices, no significant difference between model predictions and manual delineations was found. Delineation time per patient could be reduced from 40 to 12 min, including manual corrections of model predictions, and to 4 min without corrections.

CONCLUSIONS

Our Pelvic U-Net led to credible and clinically applicable OAR segmentations and showed improved performance compared to previous studies. Even though manual adjustments were needed for some predicted structures, segmentation time could be reduced by 70% on average. This allows for an accelerated radiation therapy treatment planning workflow for anal cancer patients.

Deep Learning Model for Classifying Metastatic Epidural Spinal Cord Compression on MRI

Background

Metastatic epidural spinal cord compression (MESCC) is a devastating complication of advanced cancer. A deep learning (DL) model for automated MESCC classification on MRI could aid earlier diagnosis and referral.

Purpose

To develop a DL model for automated classification of MESCC on MRI.

Materials and Methods

Patients with known MESCC diagnosed on MRI between September 2007 and September 2017 were eligible. MRI studies with instrumentation, suboptimal image quality, and non-thoracic regions were excluded. Axial T2-weighted images were utilized. The internal dataset split was 82% and 18% for training/validation and test sets, respectively. External testing was also performed. Internal training/validation data were labeled using the Bilsky MESCC classification by a musculoskeletal radiologist (10-year experience) and a neuroradiologist (5-year experience). These labels were used to train a DL model utilizing a prototypical convolutional neural network. Internal and external test sets were labeled by the musculoskeletal radiologist as the reference standard. For assessment of DL model performance and interobserver variability, test sets were labeled independently by the neuroradiologist (5-year experience), a spine surgeon (5-year experience), and a radiation oncologist (11-year experience). Inter-rater agreement (Gwet’s kappa) and sensitivity/specificity were calculated.

Results

Overall, 215 MRI spine studies were analyzed [164 patients, mean age = 62 ± 12(SD)] with 177 (82%) for training/validation and 38 (18%) for internal testing. For internal testing, the DL model and specialists all showed almost perfect agreement (kappas = 0.92–0.98, p < 0.001) for dichotomous Bilsky classification (low versus high grade) compared to the reference standard. Similar performance was seen for external testing on a set of 32 MRI spines with the DL model and specialists all showing almost perfect agreement (kappas = 0.94–0.95, p < 0.001) compared to the reference standard.

Conclusion

A DL model showed comparable agreement to a subspecialist radiologist and clinical specialists for the classification of malignant epidural spinal cord compression and could optimize earlier diagnosis and surgical referral.