Accelerated IVIM-corrected DTI in acute hamstring injury: towards a clinically feasible acquisition time.

BACKGROUND: Intravoxel incoherent motion (IVIM)-corrected diffusion tensor imaging (DTI) potentially enhances return-to-play (RTP) prediction after hamstring injuries. However, the long scan times hamper clinical implementation. We assessed accelerated IVIM-corrected DTI approaches in acute hamstring injuries and explore the sensitivity of the perfusion fraction (f) to acute muscle damage. METHODS: Athletes with acute hamstring injury received DTI scans of both thighs < 7 days after injury and at RTP. For a subset, DTI scans were repeated with multiband (MB) acceleration. Data from standard and MB-accelerated scans were fitted with standard and accelerated IVIM-corrected DTI approach using high b-values only. Segmentations of the injury and contralateral healthy muscles were contoured. The fitting methods as well as the standard and MB-accelerated scan were compared using linear regression analysis. For sensitivity to injury, Δ(injured minus healthy) DTI parameters between the methods and the differences between injured and healthy muscles were compared (Wilcoxon signed-rank test). RESULTS: The baseline dataset consisted of 109 athletes (16 with MB acceleration); 64 of them received an RTP scan (8 with MB acceleration). Linear regression of the standard and high-b DTI fitting showed excellent agreement. With both fitting methods, standard and MB-accelerated scans were comparable. Δ(injured minus healthy) was similar between standard and accelerated methods. For all methods, all IVIM-DTI parameters except f were significantly different between injured and healthy muscles. CONCLUSIONS: High-b DTI fitting with MB acceleration reduced the scan time from 11:08 to 3:40 min:s while maintaining sensitivity to hamstring injuries; f was not different between healthy and injured muscles. RELEVANCE STATEMENT: The accelerated IVIM-corrected DTI protocol, using fewer b-values and MB acceleration, reduced the scan time to under 4 min without affecting the sensitivity of the quantitative outcome parameters to hamstring injuries. This allows for routine clinical monitoring of hamstring injuries, which could directly benefit injury treatment and monitoring. KEY POINTS: • Combining high-b DTI-fitting and multiband-acceleration dramatically reduced by two thirds the scan time. • The accelerated IVIM-corrected DTI approaches maintained the sensitivity to hamstring injuries. • The IVIM-derived perfusion fraction was not sensitive to hamstring injuries.

Contrast-agent-based perfusion MRI code repository and testing framework: ISMRM Open Science Initiative for Perfusion Imaging (OSIPI).

PURPOSE: Software has a substantial impact on quantitative perfusion MRI values. The lack of generally accepted implementations, code sharing and transparent testing reduces reproducibility, hindering the use of perfusion MRI in clinical trials. To address these issues, the ISMRM Open Science Initiative for Perfusion Imaging (OSIPI) aimed to establish a community-led, centralized repository for sharing open-source code for processing contrast-based perfusion imaging, incorporating an open-source testing framework. METHODS: A repository was established on the OSIPI GitHub website. Python was chosen as the target software language. Calls for code contributions were made to OSIPI members, the ISMRM Perfusion Study Group, and publicly via OSIPI websites. An automated unit-testing framework was implemented to evaluate the output of code contributions, including visual representation of the results. RESULTS: The repository hosts 86 implementations of perfusion processing steps contributed by 12 individuals or teams. These cover all core aspects of DCE- and DSC-MRI processing, including multiple implementations of the same functionality. Tests were developed for 52 implementations, covering five analysis steps. For T1 mapping, signal-to-concentration conversion and population AIF functions, different implementations resulted in near-identical output values. For the five pharmacokinetic models tested (Tofts, extended Tofts-Kety, Patlak, two-compartment exchange, and two-compartment uptake), differences in output parameters were observed between contributions. CONCLUSIONS: The OSIPI DCE-DSC code repository represents a novel community-led model for code sharing and testing. The repository facilitates the re-use of existing code and the benchmarking of new code, promoting enhanced reproducibility in quantitative perfusion imaging.

The ISMRM Open Science Initiative for Perfusion Imaging (OSIPI): Results from the OSIPI-Dynamic Contrast-Enhanced challenge.

PURPOSE: K trans $$ {K}^{\mathrm{trans}} $$ has often been proposed as a quantitative imaging biomarker for diagnosis, prognosis, and treatment response assessment for various tumors. None of the many software tools for K trans $$ {K}^{\mathrm{trans}} $$ quantification are standardized. The ISMRM Open Science Initiative for Perfusion Imaging-Dynamic Contrast-Enhanced (OSIPI-DCE) challenge was designed to benchmark methods to better help the efforts to standardize K trans $$ {K}^{\mathrm{trans}} $$ measurement. METHODS: A framework was created to evaluate K trans $$ {K}^{\mathrm{trans}} $$ values produced by DCE-MRI analysis pipelines to enable benchmarking. The perfusion MRI community was invited to apply their pipelines for K trans $$ {K}^{\mathrm{trans}} $$ quantification in glioblastoma from clinical and synthetic patients. Submissions were required to include the entrants' K trans $$ {K}^{\mathrm{trans}} $$ values, the applied software, and a standard operating procedure. These were evaluated using the proposed OSIP I gold $$ \mathrm{OSIP}{\mathrm{I}}_{\mathrm{gold}} $$ score defined with accuracy, repeatability, and reproducibility components. RESULTS: Across the 10 received submissions, the OSIP I gold $$ \mathrm{OSIP}{\mathrm{I}}_{\mathrm{gold}} $$ score ranged from 28% to 78% with a 59% median. The accuracy, repeatability, and reproducibility scores ranged from 0.54 to 0.92, 0.64 to 0.86, and 0.65 to 1.00, respectively (0-1 = lowest-highest). Manual arterial input function selection markedly affected the reproducibility and showed greater variability in K trans $$ {K}^{\mathrm{trans}} $$ analysis than automated methods. Furthermore, provision of a detailed standard operating procedure was critical for higher reproducibility. CONCLUSIONS: This study reports results from the OSIPI-DCE challenge and highlights the high inter-software variability within K trans $$ {K}^{\mathrm{trans}} $$ estimation, providing a framework for ongoing benchmarking against the scores presented. Through this challenge, the participating teams were ranked based on the performance of their software tools in the particular setting of this challenge. In a real-world clinical setting, many of these tools may perform differently with different benchmarking methodology.

Improving microstructural integrity, interstitial fluid, and blood microcirculation images from multi-b-value diffusion MRI using physics-informed neural networks in cerebrovascular disease.

PURPOSE: To obtain better microstructural integrity, interstitial fluid, and microvascular images from multi-b-value diffusion MRI data by using a physics-informed neural network (PINN) fitting approach. METHODS: Test-retest whole-brain inversion recovery diffusion-weighted images with multiple b-values (IVIM: intravoxel incoherent motion) were acquired on separate days for 16 patients with cerebrovascular disease on a 3.0T MRI system. The performance of the PINN three-component IVIM (3C-IVIM) model fitting approach was compared with conventional fitting approaches (i.e., non-negative least squares and two-step least squares) in terms of (1) parameter map quality, (2) test-retest repeatability, and (3) voxel-wise accuracy. Using the in vivo data, the parameter map quality was assessed by the parameter contrast-to-noise ratio (PCNR) between normal-appearing white matter and white matter hyperintensities, and test-retest repeatability was expressed by the coefficient of variation (CV) and intraclass correlation coefficient (ICC). The voxel-wise accuracy of the 3C-IVIM parameters was determined by 10,000 computer simulations mimicking our in vivo data. Differences in PCNR and CV values obtained with the PINN approach versus conventional fitting approaches were assessed using paired Wilcoxon signed-rank tests. RESULTS: The PINN-derived 3C-IVIM parameter maps were of higher quality and more repeatable than those of conventional fitting approaches, while also achieving higher voxel-wise accuracy. CONCLUSION: Physics-informed neural networks enable robust voxel-wise estimation of three diffusion components from the diffusion-weighted signal. The repeatable and high-quality biological parameter maps generated with PINNs allow for visual evaluation of pathophysiological processes in cerebrovascular disease.

Model-based reconstructions for intravoxel incoherent motion and diffusion tensor imaging parameter map estimations.

Intravoxel incoherent motion (IVIM) imaging and diffusion tensor imaging (DTI) facilitate noninvasive quantification of tissue perfusion and diffusion. Both are promising biomarkers in various diseases and a combined acquisition is therefore desirable. This comes with challenges, including noisy parameter maps and long scan times, especially for the perfusion fraction f and pseudo-diffusion coefficient D*. A model-based reconstruction has the potential to overcome these challenges. As a first step, our goal was to develop a model-based reconstruction framework for IVIM and combined IVIM-DTI parameter estimation. The IVIM and IVIM-DTI models were implemented in the PyQMRI model-based reconstruction framework and validated with simulations and in vivo data. Commonly used voxel-wise nonlinear least-squares fitting was used as the reference. Simulations with the IVIM and IVIM-DTI models were performed with 100 noise realizations to assess accuracy and precision. Diffusion-weighted data were acquired for IVIM reconstruction in the liver (n = 5), as well as for IVIM-DTI in the kidneys (n = 5) and lower-leg muscles (n = 6) of healthy volunteers. The median and interquartile range (IQR) values of the IVIM and IVIM-DTI parameters were compared to assess bias and precision. With model-based reconstruction, the parameter maps exhibited less noise, which was most pronounced in the f and D* maps, both in the simulations and in vivo. The bias values in the simulations were comparable between model-based reconstruction and the reference method. The IQR was lower with model-based reconstruction compared with the reference for all parameters. In conclusion, model-based reconstruction is feasible for IVIM and IVIM-DTI and improves the precision of the parameter estimates, particularly for f and D* maps.

Deep Learning for Automated Elective Lymph Node Level Segmentation for Head and Neck Cancer Radiotherapy.

Depending on the clinical situation, different combinations of lymph node (LN) levels define the elective LN target volume in head-and-neck cancer (HNC) radiotherapy. The accurate auto-contouring of individual LN levels could reduce the burden and variability of manual segmentation and be used regardless of the primary tumor location. We evaluated three deep learning approaches for the segmenting individual LN levels I−V, which were manually contoured on CT scans from 70 HNC patients. The networks were trained and evaluated using five-fold cross-validation and ensemble learning for 60 patients with (1) 3D patch-based UNets, (2) multi-view (MV) voxel classification networks and (3) sequential UNet+MV. The performances were evaluated using Dice similarity coefficients (DSC) for automated and manual segmentations for individual levels, and the planning target volumes were extrapolated from the combined levels I−V and II−IV, both for the cross-validation and for an independent test set of 10 patients. The median DSC were 0.80, 0.66 and 0.82 for UNet, MV and UNet+MV, respectively. Overall, UNet+MV significantly (p < 0.0001) outperformed other arrangements and yielded DSC = 0.87, 0.85, 0.86, 0.82, 0.77, 0.77 for the combined and individual level I−V structures, respectively. Both PTVs were also significantly (p < 0.0001) more accurate with UNet+MV, with DSC = 0.91 and 0.90, respectively. The accurate segmentation of individual LN levels I−V can be achieved using an ensemble of UNets. UNet+MV can further refine this result.

Added value of 3T MRI and the MRI-halo sign in assessing resectability of locally advanced pancreatic cancer following induction chemotherapy (IMAGE-MRI): prospective pilot study.

BACKGROUND: Restaging of locally advanced pancreatic cancer (LAPC) after induction chemotherapy using contrast-enhanced computed tomography (CE-CT) imaging is imprecise in evaluating local tumor response. This study explored the value of 3 Tesla (3 T) contrast-enhanced (CE) and diffusion-weighted (DWI) magnetic resonance imaging (MRI) for local tumor restaging. METHODS: This is a prospective pilot study including 20 consecutive patients with LAPC with RECIST non-progressive disease on CE-CT after induction chemotherapy. Restaging CE-CT, CE-MRI, and DWI-MRI were retrospectively evaluated by two abdominal radiologists in consensus, scoring tumor size and vascular involvement. A halo sign was defined as replacement of solid perivascular (arterial and venous) tumor tissue by a zone of fatty-like signal intensity. RESULTS: Adequate MRI was obtained in 19 patients with LAPC after induction chemotherapy. Tumor diameter was non-significantly smaller on CE-MRI compared to CE-CT (26 mm vs. 30 mm; p = 0.073). An MRI-halo sign was seen on CE-MRI in 52.6% (n = 10/19), whereas a CT-halo sign was seen in 10.5% (n = 2/19) of patients (p = 0.016). An MRI-halo sign was not associated with resection rate (60.0% vs. 62.5%; p = 1.000). In the resection cohort, patients with an MRI-halo sign had a non-significant increased R0 resection rate as compared to patients without an MRI-halo sign (66.7% vs. 20.0%; p = 0.242). Positive and negative predictive values of the CE-MRI-halo sign for R0 resection were 66.7% and 66.7%, respectively. CONCLUSIONS: 3 T CE-MRI and the MRI-halo sign might be helpful to assess the effect of induction chemotherapy in patients with LAPC, but its diagnostic accuracy has to be evaluated in larger series.

Potential of Deep Learning in Quantitative Magnetic Resonance Imaging for Personalized Radiotherapy.

Quantitative magnetic resonance imaging (qMRI) has been shown to provide many potential advantages for personalized adaptive radiotherapy (RT). Deep learning models have proven to increase efficiency, robustness and speed for different qMRI tasks. Therefore, this article discusses the current state-of-the-art and potential future opportunities as well as challenges related to the use of deep learning in qMRI for target contouring, quantitative parameter estimation and also the generation of synthetic computerized tomography (CT) data based on MRI in personalized RT.

The future of MRI in radiation therapy: Challenges and opportunities for the MR community.

Radiation therapy is a major component of cancer treatment pathways worldwide. The main aim of this treatment is to achieve tumor control through the delivery of ionizing radiation while preserving healthy tissues for minimal radiation toxicity. Because radiation therapy relies on accurate localization of the target and surrounding tissues, imaging plays a crucial role throughout the treatment chain. In the treatment planning phase, radiological images are essential for defining target volumes and organs-at-risk, as well as providing elemental composition (e.g., electron density) information for radiation dose calculations. At treatment, onboard imaging informs patient setup and could be used to guide radiation dose placement for sites affected by motion. Imaging is also an important tool for treatment response assessment and treatment plan adaptation. MRI, with its excellent soft tissue contrast and capacity to probe functional tissue properties, holds great untapped potential for transforming treatment paradigms in radiation therapy. The MR in Radiation Therapy ISMRM Study Group was established to provide a forum within the MR community to discuss the unmet needs and fuel opportunities for further advancement of MRI for radiation therapy applications. During the summer of 2021, the study group organized its first virtual workshop, attended by a diverse international group of clinicians, scientists, and clinical physicists, to explore our predictions for the future of MRI in radiation therapy for the next 25 years. This article reviews the main findings from the event and considers the opportunities and challenges of reaching our vision for the future in this expanding field.

Self-supervised neural network improves tri-exponential intravoxel incoherent motion model fitting compared to least-squares fitting in non-alcoholic fatty liver disease.

Recent literature suggests that tri-exponential models may provide additional information and fit liver intravoxel incoherent motion (IVIM) data more accurately than conventional bi-exponential models. However, voxel-wise fitting of IVIM results in noisy and unreliable parameter maps. For bi-exponential IVIM, neural networks (NN) were able to produce superior parameter maps than conventional least-squares (LSQ) generated images. Hence, to improve parameter map quality of tri-exponential IVIM, we developed an unsupervised physics-informed deep neural network (IVIM3-NET). We assessed its performance in simulations and in patients with non-alcoholic fatty liver disease (NAFLD) and compared outcomes with bi-exponential LSQ and NN fits and tri-exponential LSQ fits. Scanning was performed using a 3.0T free-breathing multi-slice diffusion-weighted single-shot echo-planar imaging sequence with 18 b-values. Images were analysed for visual quality, comparing the bi- and tri-exponential IVIM models for LSQ fits and NN fits using parameter-map signal-to-noise ratios (SNR) and adjusted R 2. IVIM parameters were compared to histological fibrosis, disease activity and steatosis grades. Parameter map quality improved with bi- and tri-exponential NN approaches, with a significant increase in average parameter-map SNR from 3.38 to 5.59 and 2.45 to 4.01 for bi- and tri-exponential LSQ and NN models respectively. In 33 out of 36 patients, the tri-exponential model exhibited higher adjusted R 2 values than the bi-exponential model. Correlating IVIM data to liver histology showed that the bi- and tri-exponential NN outperformed both LSQ models for the majority of IVIM parameters (10 out of 15 significant correlations). Overall, our results support the use of a tri-exponential IVIM model in NAFLD. We show that the IVIM3-NET can be used to improve image quality compared to a tri-exponential LSQ fit and provides promising correlations with histopathology similar to the bi-exponential neural network fit, while generating potentially complementary additional parameters.

Deep learning DCE-MRI parameter estimation: Application in pancreatic cancer.

Dynamic contrast-enhanced magnetic resonance imaging (DCE-MRI) is an MRI technique for quantifying perfusion that can be used in clinical applications for classification of tumours and other types of diseases. Conventionally, the non-linear least squares (NLLS) methods is used for tracer-kinetic modelling of DCE data. However, despite promising results, NLLS suffers from long processing times (minutes-hours) and noisy parameter maps due to the non-convexity of the cost function. In this work, we investigated physics-informed deep neural networks for estimating physiological parameters from DCE-MRI signal-curves. Three voxel-wise temporal frameworks (FCN, LSTM, GRU) and two spatio-temporal frameworks (CNN, U-Net) were investigated. The accuracy and precision of parameter estimation by the temporal frameworks were evaluated in simulations. All networks showed higher precision than the NLLS. Specifically, the GRU showed to decrease the random error on ve by a factor of 4.8 with respect to the NLLS for noise (SD) of 1/20. The accuracy was better for the prediction of the ve parameter in all networks compared to the NLLS. The GRU and LSTM worked with arbitrary acquisition lengths. The GRU was selected for in vivo evaluation and compared to the spatio-temporal frameworks in 28 patients with pancreatic cancer. All neural network approaches showed less noisy parameter maps than the NLLS. The GRU had better test-retest repeatability than the NLLS for all three parameters and was able to detect one additional patient with significant changes in DCE parameters post chemo-radiotherapy. Although the U-Net and CNN had even better test-retest characteristics than the GRU, and were able to detect even more responders, they also showed potential systematic errors in the parameter maps. Therefore, we advise using our GRU framework for analysing DCE data.

Dynamic MRI of swallowing: real-time volumetric imaging at 12 frames per second at 3 T.

OBJECTIVE: Dysphagia or difficulty in swallowing is a potentially hazardous clinical problem that needs regular monitoring. Real-time 2D MRI of swallowing is a promising radiation-free alternative to the current clinical standard: videofluoroscopy. However, aspiration may be missed if it occurs outside this single imaged slice. We therefore aimed to image swallowing in 3D real time at 12 frames per second (fps). MATERIALS AND METHODS: At 3 T, three 3D real-time MRI acquisition approaches were compared to the 2D acquisition: an aligned stack-of-stars (SOS), and a rotated SOS with a golden-angle increment and with a tiny golden-angle increment. The optimal 3D acquisition was determined by computer simulations and phantom scans. Subsequently, five healthy volunteers were scanned and swallowing parameters were measured. RESULTS: Although the rotated SOS approaches resulted in better image quality in simulations, in practice, the aligned SOS performed best due to the limited number of slices. The four swallowing phases could be distinguished in 3D real-time MRI, even though the spatial blurring was stronger than in 2D. The swallowing parameters were similar between 2 and 3D. CONCLUSION: At a spatial resolution of 2-by-2-by-6 mm with seven slices, swallowing can be imaged in 3D real time at a frame rate of 12 fps.

Phase I/II Study of LDE225 in Combination with Gemcitabine and Nab-Paclitaxel in Patients with Metastatic Pancreatic Cancer.

BACKGROUND: Desmoplasia is a central feature of the tumor microenvironment in pancreatic ductal adenocarcinoma (PDAC). LDE225 is a pharmacological Hedgehog signaling pathway inhibitor and is thought to specifically target tumor stroma. We investigated the combined use of LDE225 and chemotherapy to treat PDAC patients. METHODS: This was a multi-center, phase I/II study for patients with metastatic PDAC establishing the maximum tolerated dose of LDE225 co-administered with gemcitabine and nab-paclitaxel (phase I) and evaluating the efficacy and safety of the treatment combination after prior FOLFIRINOX treatment (phase II). Tumor microenvironment assessment was performed with quantitative MRI using intra-voxel incoherent motion diffusion weighted MRI (IVIM-DWI) and dynamic contrast-enhanced (DCE) MRI. RESULTS: The MTD of LDE225 was 200 mg once daily co-administered with gemcitabine 1000 mg/m2 and nab-paclitaxel 125 mg/m2. In phase II, six therapy-related grade 4 adverse events (AE) and three grade 5 were observed. In 24 patients, the target lesion response was evaluable. Three patients had partial response (13%), 14 patients showed stable disease (58%), and 7 patients had progressive disease (29%). Median overall survival (OS) was 6 months (IQR 3.9-8.1). Blood plasma fraction (DCE) and diffusion coefficient (IVIM-DWI) significantly increased during treatment. Baseline perfusion fraction could predict OS (>222 days) with 80% sensitivity and 85% specificity. CONCLUSION: LDE225 in combination with gemcitabine and nab-paclitaxel was well-tolerated in patients with metastatic PDAC and has promising efficacy after prior treatment with FOLFIRINOX. Quantitative MRI suggested that LDE225 causes increased tumor diffusion and works particularly well in patients with poor baseline tumor perfusion.

Improved unsupervised physics-informed deep learning for intravoxel incoherent motion modeling and evaluation in pancreatic cancer patients.

PURPOSE: Earlier work showed that IVIM-NETorig , an unsupervised physics-informed deep neural network, was faster and more accurate than other state-of-the-art intravoxel-incoherent motion (IVIM) fitting approaches to diffusion-weighted imaging (DWI). This study presents a substantially improved version, IVIM-NEToptim , and characterizes its superior performance in pancreatic cancer patients. METHOD: In simulations (signal-to-noise ratio [SNR] = 20), the accuracy, independence, and consistency of IVIM-NET were evaluated for combinations of hyperparameters (fit S0, constraints, network architecture, number of hidden layers, dropout, batch normalization, learning rate), by calculating the normalized root-mean-square error (NRMSE), Spearman's ρ, and the coefficient of variation (CVNET ), respectively. The best performing network, IVIM-NEToptim was compared to least squares (LS) and a Bayesian approach at different SNRs. IVIM-NEToptim 's performance was evaluated in an independent dataset of 23 patients with pancreatic ductal adenocarcinoma. Fourteen of the patients received no treatment between two repeated scan sessions and nine received chemoradiotherapy between the repeated sessions. Intersession within-subject standard deviations (wSD) and treatment-induced changes were assessed. RESULTS: In simulations (SNR = 20), IVIM-NEToptim outperformed IVIM-NETorig in accuracy (NRMSE(D) = 0.177 vs 0.196; NMRSE(f) = 0.220 vs 0.267; NMRSE(D*) = 0.386 vs 0.393), independence (ρ(D*, f) = 0.22 vs 0.74), and consistency (CVNET (D) = 0.013 vs 0.104; CVNET (f) = 0.020 vs 0.054; CVNET (D*) = 0.036 vs 0.110). IVIM-NEToptim showed superior performance to the LS and Bayesian approaches at SNRs < 50. In vivo, IVIM-NEToptim showed significantly less noisy parameter maps with lower wSD for D and f than the alternatives. In the treated cohort, IVIM-NEToptim detected the most individual patients with significant parameter changes compared to day-to-day variations. CONCLUSION: IVIM-NEToptim is recommended for accurate, informative, and consistent IVIM fitting to DWI data.

Rapid 4D-MRI reconstruction using a deep radial convolutional neural network: Dracula.

BACKGROUND AND PURPOSE: 4D and midposition MRI could inform plan adaptation in lung and abdominal MR-guided radiotherapy. We present deep learning-based solutions to overcome long 4D-MRI reconstruction times while maintaining high image quality and short scan times. METHODS: Two 3D U-net deep convolutional neural networks were trained to accelerate the 4D joint MoCo-HDTV reconstruction. For the first network, gridded and joint MoCo-HDTV-reconstructed 4D-MRI were used as input and target data, respectively, whereas the second network was trained to directly calculate the midposition image. For both networks, input and target data had dimensions of 256 × 256 voxels (2D) and 16 respiratory phases. Deep learning-based MRI were verified against joint MoCo-HDTV-reconstructed MRI using the structural similarity index (SSIM) and the naturalness image quality evaluator (NIQE). Moreover, two experienced observers contoured the gross tumour volume and scored the images in a blinded study. RESULTS: For 12 subjects, previously unseen by the networks, high-quality 4D and midposition MRI (1.25 × 1.25 × 3.3 mm3) were each reconstructed from gridded images in only 28 seconds per subject. Excellent agreement was found between deep-learning-based and joint MoCo-HDTV-reconstructed MRI (average SSIM ≥ 0.96, NIQE scores 7.94 and 5.66). Deep-learning-based 4D-MRI were clinically acceptable for target and organ-at-risk delineation. Tumour positions agreed within 0.7 mm on midposition images. CONCLUSION: Our results suggest that the joint MoCo-HDTV and midposition algorithms can each be approximated by a deep convolutional neural network. This rapid reconstruction of 4D and midposition MRI facilitates online treatment adaptation in thoracic or abdominal MR-guided radiotherapy.

Repeatability of IVIM biomarkers from diffusion-weighted MRI in head and neck: Bayesian probability versus neural network.

PURPOSE: The intravoxel incoherent motion (IVIM) model for DWI might provide useful biomarkers for disease management in head and neck cancer. This study compared the repeatability of three IVIM fitting methods to the conventional nonlinear least-squares regression: Bayesian probability estimation, a recently introduced neural network approach, IVIM-NET, and a version of the neural network modified to increase consistency, IVIM-NETmod . METHODS: Ten healthy volunteers underwent two imaging sessions of the neck, two weeks apart, with two DWI acquisitions per session. Model parameters (ADC, diffusion coefficient Dt , perfusion fraction fp , and pseudo-diffusion coefficient Dp ) from each fit method were determined in the tonsils and in the pterygoid muscles. Within-subject coefficients of variation (wCV) were calculated to assess repeatability. Training of the neural network was repeated 100 times with random initialization to investigate consistency, quantified by the coefficient of variance. RESULTS: The Bayesian and neural network approaches outperformed nonlinear regression in terms of wCV. Intersession wCV of Dt in the tonsils was 23.4% for nonlinear regression, 9.7% for Bayesian estimation, 9.4% for IVIM-NET, and 11.2% for IVIM-NETmod . However, results from repeated training of the neural network on the same data set showed differences in parameter estimates: The coefficient of variances over the 100 repetitions for IVIM-NET were 15% for both Dt and fp , and 94% for Dp ; for IVIM-NETmod , these values improved to 5%, 9%, and 62%, respectively. CONCLUSION: Repeatabilities from the Bayesian and neural network approaches are superior to that of nonlinear regression for estimating IVIM parameters in the head and neck.

Cross-modality deep learning: Contouring of MRI data from annotated CT data only.

PURPOSE: Online adaptive radiotherapy would greatly benefit from the development of reliable auto-segmentation algorithms for organs-at-risk and radiation targets. Current practice of manual segmentation is subjective and time-consuming. While deep learning-based algorithms offer ample opportunities to solve this problem, they typically require large datasets. However, medical imaging data are generally sparse, in particular annotated MR images for radiotherapy. In this study, we developed a method to exploit the wealth of publicly available, annotated CT images to generate synthetic MR images, which could then be used to train a convolutional neural network (CNN) to segment the parotid glands on MR images of head and neck cancer patients. METHODS: Imaging data comprised 202 annotated CT and 27 annotated MR images. The unpaired CT and MR images were fed into a 2D CycleGAN network to generate synthetic MR images from the CT images. Annotations of axial slices of the synthetic images were generated by propagating the CT contours. These were then used to train a 2D CNN. We assessed the segmentation accuracy using the real MR images as test dataset. The accuracy was quantified with the 3D Dice similarity coefficient (DSC), Hausdorff distance (HD), and mean surface distance (MSD) between manual and auto-generated contours. We benchmarked the approach by a comparison to the interobserver variation determined for the real MR images, as well as to the accuracy when training the 2D CNN to segment the CT images. RESULTS: The determined accuracy (DSC: 0.77±0.07, HD: 18.04±12.59mm, MSD: 2.51±1.47mm) was close to the interobserver variation (DSC: 0.84±0.06, HD: 10.85±5.74mm, MSD: 1.50±0.77mm), as well as to the accuracy when training the 2D CNN to segment the CT images (DSC: 0.81±0.07, HD: 13.00±7.61mm, MSD: 1.87±0.84mm). CONCLUSIONS: The introduced cross-modality learning technique can be of great value for segmentation problems with sparse training data. We anticipate using this method with any nonannotated MRI dataset to generate annotated synthetic MR images of the same type via image style transfer from annotated CT images. Furthermore, as this technique allows for fast adaptation of annotated datasets from one imaging modality to another, it could prove useful for translating between large varieties of MRI contrasts due to differences in imaging protocols within and between institutions.

A convolutional neural network for contouring metastatic lymph nodes on diffusion-weighted magnetic resonance images for assessment of radiotherapy response.

BACKGROUND AND PURPOSE: Retrieving quantitative parameters from magnetic resonance imaging (MRI), e.g. for early assessment of radiotherapy treatment response, necessitates contouring regions of interest, which is time-consuming and prone to errors. This becomes more pressing for daily imaging on MRI-guided radiotherapy systems. Therefore, we trained a deep convolutional neural network to automatically contour involved lymph nodes on diffusion-weighted (DW) MRI of head and neck cancer (HNC) patients receiving radiotherapy. MATERIALS AND METHODS: DW-images from 48 HNC patients (18 induction-chemotherapy + chemoradiotherapy; 30 definitive chemoradiotherapy) with 68 involved lymph nodes were obtained on a diagnostic 1.5 T MR-scanner prior to and 2-3 timepoints throughout treatment. A radiation oncologist delineated the lymph nodes on the b = 50 s/mm2 images. A 3D U-net was trained to contour involved lymph nodes. Its performance was evaluated in all 48 patients using 8-fold cross-validation and calculating the Dice similarity coefficient (DSC) and the absolute difference in median apparent diffusion coefficient (ΔADC) between the manual and generated contours. Additionally, the performance was evaluated in an independent dataset of three patients obtained on a 1.5 T MR-Linac. RESULTS: In the definitive chemoradiotherapy patients (n = 96 patients/lymphnodes/timepoints) the DSC was 0.87 (0.81-0.91) [median (1st-3rd quantiles)] and ΔADC was 1.9% (0.8-3.4%) and both remained stable throughout treatment. The network performed worse in the patients receiving induction-chemotherapy (n = 65), with DSC = 0.80 (0.71-0.87) and ΔADC = 3.3% (1.6-8.0%). The network performed well on the MR-Linac data (n = 8) with DSC = 0.80 (0.75-0.82) and ΔADC = 4.0% (0.6-9.1%). CONCLUSIONS: We established accurate automatic contouring of involved lymph nodes for HNC patients on diagnostic and MR-Linac DW-images.

Pathological validation and prognostic potential of quantitative MRI in the characterization of pancreas cancer: preliminary experience.

Patient stratification based on biological variation in pancreatic ductal adenocarcinoma (PDAC) subtypes could help to improve clinical outcome. However, noninvasive assessment of the entire tumor microenvironment remains challenging. In this study, we investigate the biological basis of dynamic contrast-enhanced (DCE), intravoxel incoherent motion (IVIM), and R2*-derived magnetic resonance imaging (MRI) parameters for the noninvasive characterization of the PDAC tumor microenvironment and evaluate their prognostic potential in PDAC patients. Patients diagnosed with treatment-naïve resectable PDAC underwent MRI. After resection, a whole-mount tumor slice was analyzed for collagen fraction, vessel density, and hypoxia and matched to the MRI parameter maps. MRI parameters were correlated to immunohistochemistry-derived tissue characteristics and evaluated for prognostic potential. Thirty patients were included of whom 21 underwent resection with whole-mount histology available in 15 patients. DCE Ktrans and ve , ADC, and IVIM D correlated with collagen fraction. DCE kep and IVIM f correlated with vessel density and R2* with tissue hypoxia. Based on MRI, two main PDAC phenotypes could be distinguished; a stroma-high phenotype demonstrating high vessel density and high collagen fraction and a stroma-low phenotype demonstrating low vessel density and low collagen fraction. Patients with the stroma-high phenotype (high kep and high IVIM D, n = 8) showed longer overall survival (not reached vs. 14 months, P = 0.001, HR = 9.1, P = 0.004) and disease-free survival (not reached vs. 2 months, P < 0.001, HR 9.3, P = 0.003) compared to the other patients (n = 22). Median follow-up was 41 (95% CI: 36-46) months. MRI was able to accurately characterize tumor collagen fraction, vessel density, and hypoxia in PDAC. Based on imaging parameters, a subgroup of patients with significantly better prognosis could be identified. These first results indicate that stratification-based MRI-derived biomarkers could help to tailor treatment and improve clinical outcome and warrant further research.

Quantitative imaging for radiotherapy purposes.

Quantitative imaging biomarkers show great potential for use in radiotherapy. Quantitative images based on microscopic tissue properties and tissue function can be used to improve contouring of the radiotherapy targets. Furthermore, quantitative imaging biomarkers might be used to predict treatment response for several treatment regimens and hence be used as a tool for treatment stratification, either to determine which treatment modality is most promising or to determine patient-specific radiation dose. Finally, patient-specific radiation doses can be further tailored to a tissue/voxel specific radiation dose when quantitative imaging is used for dose painting. In this review, published standards, guidelines and recommendations on quantitative imaging assessment using CT, PET and MRI are discussed. Furthermore, critical issues regarding the use of quantitative imaging for radiation oncology purposes and resultant pending research topics are identified.