Quantifying aortic valve regurgitation in patients with congenital aortic valve disease by 2D and 4D flow magnetic resonance analysis.

BACKGROUND: In congenital aortic valve disease, quantifying aortic regurgitation (AR) varies by the measurement site. Our study aimed to identify the optimal site for AR assessment using 2D and 4D MR flow measurements, with a focus on vortices. METHODS: We retrospectively analysed 31 patients with congenital aortic valve disease, performing 2D and 4D MR flow measurements at the aortic valve, sinotubular junction (STJ), ascending aorta (AAo), and using midpulmonary artery measurements as a reference. We assessed percentage AR and net forward volumes, calculated linear correlations, and plotted Bland-Altman plots. Net forward flow at all aortic sites were correlated with the main pulmonary artery. Differences in AR between 2D and 4D flows were linked to vortices detected by 4D streamlines. RESULTS: The best agreement in % AR between 2D and 4D flows was at the aortic valve (mean difference 4D2D -2.9%, limits of agreement 8.7% to -14.3%; r2 = 0.7). Correlations weakened at STJ and AAo. Vortices in the ascending aorta led to AR overestimation in 2D measurements. Net forward flow at the aortic valve by 4D flow correlated closer with main pulmonary artery than did 2D flow. (Mean difference for 2D and 4D MR flow 7.5 ml and 4.2 ml, respectively). CONCLUSIONS: For congenital aortic valve disease, the most accurate AR quantification occurs at the aortic valve using 2D and 4D MR flow. Notably, vortices in the ascending aorta can result in AR overestimation with 2D MR flow.

Interrelation Between Cerebrospinal Fluid Pressure, Intracranial Morphology and Venous Hemodynamics Studied by 4D Flow MRI.

PURPOSE: To quantify the effects of CSF pressure alterations on intracranial venous morphology and hemodynamics in idiopathic intracranial hypertension (IIH) and spontaneous intracranial hypotension (SIH) and assess reversibility when the underlying cause is resolved. METHODS: We prospectively examined venous volume, intracranial venous blood flow and velocity, including optic nerve sheath diameter (ONSD) as a noninvasive surrogate of CSF pressure changes in 11 patients with IIH, 11 age-matched and sex-matched healthy controls and 9 SIH patients, before and after neurosurgical closure of spinal dural leaks. We applied multiparametric MRI including 4D flow MRI, time-of-flight (TOF) and T2-weighted half-Fourier acquisition single-shot turbo-spin echo (HASTE). RESULTS: Sinus volume overlapped between groups at baseline but decreased after treatment of intracranial hypotension (p = 0.067) along with a significant increase of ONSD (p = 0.003). Blood flow in the middle and dorsal superior sagittal sinus was remarkably lower in patients with higher CSF pressure (i.e., IIH versus controls and SIH after CSF leak closure) but blood flow velocity was comparable cross-sectionally between groups and longitudinally in SIH. CONCLUSION: We were able to demonstrate the interaction of CSF pressure, venous volumetry, venous hemodynamics and ONSD using multiparametric brain MRI. Closure of CSF leaks in SIH patients resulted in symptoms suggestive of increased intracranial pressure and caused a subsequent decrease of intracranial venous volume and of blood flow within the superior sagittal sinus while ONSD increased. In contrast, blood flow parameters from 4D flow MRI did not discriminate IIH, SIH and controls as hemodynamics at baseline overlapped at most vessel cross-sections.

A simulation-based phantom model for generating synthetic mitral valve image data-application to MRI acquisition planning.

PURPOSE: Numerical phantom methods are widely used in the development of medical imaging methods. They enable quantitative evaluation and direct comparison with controlled and known ground truth information. Cardiac magnetic resonance has the potential for a comprehensive evaluation of the mitral valve (MV). The goal of this work is the development of a numerical simulation framework that supports the investigation of MRI imaging strategies for the mitral valve. METHODS: We present a pipeline for synthetic image generation based on the combination of individual anatomical 3D models with a position-based dynamics simulation of the mitral valve closure. The corresponding images are generated using modality-specific intensity models and spatiotemporal sampling concepts. We test the applicability in the context of MRI imaging strategies for the assessment of the mitral valve. Synthetic images are generated with different strategies regarding image orientation (SAX and rLAX) and spatial sampling density. RESULTS: The suitability of the imaging strategy is evaluated by comparing MV segmentations against ground truth annotations. The generated synthetic images were compared to ones acquired with similar parameters, and the result is promising. The quantitative analysis of annotation results suggests that the rLAX sampling strategy is preferable for MV assessment, reaching accuracy values that are comparable to or even outperform literature values. CONCLUSION: The proposed approach provides a valuable tool for the evaluation and optimization of cardiac valve image acquisition. Its application to the use case identifies the radial image sampling strategy as the most suitable for MV assessment through MRI.

4D Flow cardiovascular magnetic resonance consensus statement: 2023 update.

Hemodynamic assessment is an integral part of the diagnosis and management of cardiovascular disease. Four-dimensional cardiovascular magnetic resonance flow imaging (4D Flow CMR) allows comprehensive and accurate assessment of flow in a single acquisition. This consensus paper is an update from the 2015 '4D Flow CMR Consensus Statement'. We elaborate on 4D Flow CMR sequence options and imaging considerations. The document aims to assist centers starting out with 4D Flow CMR of the heart and great vessels with advice on acquisition parameters, post-processing workflows and integration into clinical practice. Furthermore, we define minimum quality assurance and validation standards for clinical centers. We also address the challenges faced in quality assurance and validation in the research setting. We also include a checklist for recommended publication standards, specifically for 4D Flow CMR. Finally, we discuss the current limitations and the future of 4D Flow CMR. This updated consensus paper will further facilitate widespread adoption of 4D Flow CMR in the clinical workflow across the globe and aid consistently high-quality publication standards.

Carotid geometry is independently associated with complicated carotid artery plaques.

Introduction: Complicated carotid artery plaques (cCAPs) are associated with an increased risk of rupture and subsequent stroke. The geometry of the carotid bifurcation determines the distribution of local hemodynamics and could thus contribute to the development and composition of these plaques. Therefore, we studied the role of carotid bifurcation geometry in the presence of cCAPs. Methods: We investigated the association of individual vessel geometry with carotid artery plaque types in the Carotid Plaque Imaging in Acute Stroke (CAPIAS) study. After excluding arteries without plaque or with insufficient MRI quality, 354 carotid arteries from 182 patients were analyzed. Individual parameters of carotid geometry [i.e., internal carotid artery (ICA)/common carotid artery (CCA) ratio, bifurcation angle, and tortuosity) were derived from time-of-flight MR images. The lesion types of carotid artery plaques were determined according to the American Heart Association classification of lesions by multi-contrast 3T-MRI. The association between carotid geometry and a cCAP was studied using logistic regression after adjusting for age, sex, wall area, and cardiovascular risk factors. Results: Low ICA/CCA ratios (OR per SD increase 0.60 [95%CI: 0.42-0.85]; p = 0.004) and low bifurcation angles (OR 0.61 [95%CI: 0.42-0.90]; p = 0.012) were significantly associated with the presence of cCAPs after adjusting for age, sex, cardiovascular risk factors, and wall area. Tortuosity had no significant association with cCAPs. Only ICA/CCA ratio remained significant in a model containing all three geometric parameters (OR per SD increase 0.65 [95%CI: 0.45-0.94]; p = 0.023). Conclusions: A steep tapering of the ICA relative to the CCA and, to a lesser extent, a low angle of the carotid bifurcation were associated with the presence of cCAPs. Our findings highlight the contribution of bifurcation geometry to plaque vulnerability. Thus, assessment of carotid geometry could be helpful in identifying patients at risk of cCAPs.

Lazy Luna: Extendible software for multilevel reader comparison in cardiovascular magnetic resonance imaging.

BACKGROUND AND OBJECTIVES: Cardiovascular Magnetic Resonance (CMR) imaging is a growing field with increasing diagnostic utility in clinical routine. Quantitative diagnostic parameters are typically calculated based on contours or points provided by readers, e.g. natural intelligences (NI) such as clinicians or researchers, and artificial intelligences (AI). As clinical applications multiply, evaluating the precision and reproducibility of quantitative parameters becomes increasingly important. Although segmentation challenges for AIs and guidelines for clinicians provide quality assessments and regulation, the methods ought to be combined and streamlined for clinical applications. The goal of the developed software, Lazy Luna (LL), is to offer a flexible evaluation tool that is readily extendible to new sequences and scientific endeavours. METHODS: An interface was designed for LL, which allows for comparing annotated CMR images. Geometric objects ensure precise calculations of metric values and clinical results regardless of whether annotations originate from AIs or NIs. A graphical user interface (GUI) is provided to make the software available to non-programmers. The GUI allows for an interactive inspection of image datasets as well as implementing tracing procedures, which follow statistical reader differences in clinical results to their origins in individual image contours. The backend software builds on a set of meta-classes, which can be extended to new imaging sequences and clinical parameters. Following an agile development procedure with clinical feedback allows for a quick implementation of new classes, figures and tables for evaluation. RESULTS: Two application cases present LL's extendibility to clinical evaluation and AI development contexts. The first concerns T1 parametric mapping images segmented by two expert readers. Quantitative result differences are traced to reveal typical segmentation dissimilarities from which these differences originate. The meta-classes are extended to this new application scenario. The second applies to the open source Late Gadolinium Enhancement (LGE) quantification challenge for AI developers "Emidec", which illustrates LL's usability as open source software. CONCLUSION: The presented software Lazy Luna allows for an automated multilevel comparison of readers as well as identifying qualitative reasons for statistical reader differences. The open source software LL can be extended to new application cases in the future.

Radiomics-based aortic flow profile characterization with 4D phase-contrast MRI.

4D PC MRI of the aorta has become a routinely available examination, and a multitude of single parameters have been suggested for the quantitative assessment of relevant flow features for clinical studies and diagnosis. However, clinically applicable assessment of complex flow patterns is still challenging. We present a concept for applying radiomics for the quantitative characterization of flow patterns in the aorta. To this end, we derive cross-sectional scalar parameter maps related to parameters suggested in literature such as throughflow, flow direction, vorticity, and normalized helicity. Derived radiomics features are selected with regard to their inter-scanner and inter-observer reproducibility, as well as their performance in the differentiation of sex-, age- and disease-related flow properties. The reproducible features were tested on user-selected examples with respect to their suitability for characterizing flow profile types. In future work, such signatures could be applied for quantitative flow assessment in clinical studies or disease phenotyping.

Characterization of aortic aging using 3D multi-parametric MRI-long-term follow-up in a population study.

We comprehensively studied morphological and functional aortic aging in a population study using modern three-dimensional MR imaging to allow future comparison in patients with diseases of the aortic valve or aorta. We followed 80 of 126 subjects of a population study (20 to 80 years of age at baseline) using the identical methodology 6.0 ± 0.5 years later. All underwent 3 T MRI of the thoracic aorta including 3D T1 weighted MRI (spatial resolution 1 mm3) for measuring aortic diameter and plaque thickness and 4D flow MRI (spatial/temporal resolution = 2 mm3/20 ms) for calculating global and regional aortic pulse wave velocity (PWV) and helicity of aortic blood flow. Mean diameter of the ascending aorta (AAo) decreased and plaque thickness increased significantly in the aortic arch (AA) and descending aorta (DAo) in females. PWV of the thoracic aorta increased (6.4 ± 1.5 to 7.0 ± 1.7 m/s and 6.8 ± 1.5 to 7.3 ± 1.8 m/s in females and males, respectively) over time. Local normalized helicity volumes (LNHV) decreased significantly in the AAo and AA (0.33 to 0.31 and 0.34 to 0.32 in females and 0.34 to 0.32 and 0.32 to 0.28 in males). By contrast, helicity increased significantly in the DAo in both genders (0.28 to 0.29 and 0.29 to 0.30, respectively). 3D MRI was able to characterize changes in aortic diameter, plaque thickness, PWV and helicity during six years in our population. Aortic aging determined by 3D multi-parametric MRI is now available for future comparisons in patients with diseases of the aortic valve or aorta.

Aortic wall shear stress in bicuspid aortic valve disease-10-year follow-up.

Background: Bicuspid aortic valve (BAV) disease leads to deviant helical flow patterns especially in the mid-ascending aorta (AAo), potentially causing wall alterations such as aortic dilation and dissection. Among others, wall shear stress (WSS) could contribute to the prediction of long-term outcome of patients with BAV. 4D flow in cardiovascular magnetic resonance (CMR) has been established as a valid method for flow visualization and WSS estimation. The aim of this study is to reevaluate flow patterns and WSS in patients with BAV 10 years after the initial evaluation. Methods: Fifteen patients (median age 34.0 years) with BAV were re-evaluated 10 years after the initial study from 2008/2009 using 4D flow by CMR. Our particular patient cohort met the same inclusion criteria as in 2008/2009, all without enlargement of the aorta or valvular impairment at that time. Flow patterns, aortic diameters, WSS and distensibility were calculated in different aortic regions of interest (ROI) with dedicated software tools. Results: Indexed aortic diameters in the descending aorta (DAo), but especially in the AAo did not change in the 10-year period. Median difference 0.05 cm/m2 (95% CI: 0.01 to 0.22; P=0.06) for AAo and median difference -0.08 cm/m2 (95% CI: -0.12 to 0.01; P=0.07) for DAo. WSS values were lower in 2018/2019 at all measured levels. Aortic distensibility decreased by median 25.6% in the AAo, while stiffness increased concordantly (median +23.6%). Conclusions: After a ten years' follow-up of patients with isolated BAV disease, indexed aortic diameters did not change in this patient cohort. WSS was lower compared to values generated 10 years earlier. Possibly a drop of WSS in BAV could serve as a marker for a benign long-term course and implementation of more conservative treatment strategies.

Repeatability and reproducibility of various 4D Flow MRI postprocessing software programs in a multi-software and multi-vendor cross-over comparison study.

BACKGROUND: Different software programs are available for the evaluation of 4D Flow cardiovascular magnetic resonance (CMR). A good agreement of the results between programs is a prerequisite for the acceptance of the method. Therefore, the goal was to compare quantitative results from a cross-over comparison in individuals examined on two scanners of different vendors analyzed with four postprocessing software packages. METHODS: Eight healthy subjects (27 ± 3 years, 3 women) were each examined on two 3T CMR systems (Ingenia, Philips Healthcare; MAGNETOM Skyra, Siemens Healthineers) with a standardized 4D Flow CMR sequence. Six manually placed aortic contours were evaluated with Caas (Pie Medical Imaging, SW-A), cvi42 (Circle Cardiovascular Imaging, SW-B), GTFlow (GyroTools, SW-C), and MevisFlow (Fraunhofer Institute MEVIS, SW-D) to analyze seven clinically used parameters including stroke volume, peak flow, peak velocity, and area as well as typically scientifically used wall shear stress values. Statistical analysis of inter- and intrareader variability, inter-software and inter-scanner comparison included calculation of absolute and relative error (ER), intraclass correlation coefficient (ICC), Bland-Altman analysis, and equivalence testing based on the assumption that inter-software differences needed to be within 80% of the range of intrareader differences. RESULTS: SW-A and SW-C were the only software programs showing agreement for stroke volume (ICC = 0.96; ER = 3 ± 8%), peak flow (ICC: 0.97; ER = -1 ± 7%), and area (ICC = 0.81; ER = 2 ± 22%). Results from SW-A/D and SW-C/D were equivalent only for area and peak flow. Other software pairs did not yield equivalent results for routinely used clinical parameters. Especially peak maximum velocity yielded poor agreement (ICC ≤ 0.4) between all software packages except SW-A/D that showed good agreement (ICC = 0.80). Inter- and intrareader consistency for clinically used parameters was best for SW-A and SW-D (ICC = 0.56-97) and worst for SW-B (ICC = -0.01-0.71). Of note, inter-scanner differences per individual tended to be smaller than inter-software differences. CONCLUSIONS: Of all tested software programs, only SW-A and SW-C can be used equivalently for determination of stroke volume, peak flow, and vessel area. Irrespective of the applied software and scanner, high intra- and interreader variability for all parameters have to be taken into account before introducing 4D Flow CMR in clinical routine. Especially in multicenter clinical trials a single image evaluation software should be applied.

Population-based reference values for 4D flow MRI derived aortic blood flow parameters.

Objective. This study assesses age-related differences of thoracic aorta blood flow profiles and provides age- and sex-specific reference values using 4D flow cardiovascular magnetic resonance (CMR) data.Approach. 126 volunteers (age 20-80 years, female 51%) underwent 4D flow CMR and 12 perpendicular analysis planes in the thoracic aorta were specified. For these planes the following parameters were evaluated: body surface area-adjusted aortic area (A'), normalized flow displacement (NFD), the degree of wall parallelism (WPD), the minimal relative cross-sectional area through which 80% of the volume flow passes (A80) and the angle between flow direction and centerline (α).Main results. Age-related differences in blood flow parameters were seen in the ascending aorta with higher values for NFD and angle and lower values for WPD and A80 in older subjects. All parameters describing blood flow patterns correlated with the cross-sectional area in the ascending aorta. No relevant sex-differences regarding blood flow profiles were found.Significance. These age- and sex-specific reference values for quantitative parameters describing blood flow within the aorta might help to study the clinical relevance of flow profiles in the future.

Effect of transcatheter edge-to-edge repair device position on diastolic hemodynamic parameters: An echocardiography-based simulation study.

Background: Transcatheter edge-to-edge repair (TEER) has developed from innovative technology to an established treatment strategy of mitral regurgitation (MR). The risk of iatrogenic mitral stenosis after TEER is, however, a critical factor in the conflict of interest between maximal reduction of MR and minimal impairment of left ventricular filling. We aim to investigate systematically the impact of device position on the post treatment hemodynamic outcome by involving the patient-specific segmentation of the diseased mitral valve. Materials and methods: Transesophageal echocardiographic image data of ten patients with severe MR (age: 57 ± 8 years, 20% female) were segmented and virtually treated with TEER at three positions by using a position based dynamics approach. Pre- and post-interventional patient geometries were preprocessed for computational fluid dynamics (CFD) and simulated at peak-diastole with patient-specific blood flow boundary conditions. Simulations were performed with boundary conditions mimicking rest and stress. The simulation results were compared with clinical data acquired for a cohort of 21 symptomatic MR patients (age: 79 ± 6 years, 43% female) treated with TEER. Results: Virtual TEER reduces the mitral valve area (MVA) from 7.5 ± 1.6 to 2.6 ± 0.6 cm2. Central device positioning resulted in a 14% smaller MVA than eccentric device positions. Furthermore, residual MVA is better predictable for central than for eccentric device positions (R 2 = 0.81 vs. R 2 = 0.49). The MVA reduction led to significantly higher maximal diastolic velocities (pre: 0.9 ± 0.2 m/s, post: 2.0 ± 0.5 m/s) and pressure gradients (pre: 1.5 ± 0.6 mmHg, post: 16.3 ± 9 mmHg) in spite of a mean flow rate reduction by 23% due to reduced MR after the treatment. On average, velocities were 12% and pressure gradients were 25% higher with devices in central compared to lateral or medial positions. Conclusion: Virtual TEER treatment combined with CFD is a promising tool for predicting individual morphometric and hemodynamic outcomes. Such a tool can potentially be used to support clinical decision making, procedure planning, and risk estimation to prevent post-procedural iatrogenic mitral stenosis.

Virtual reality in cardiac interventions-New tools or new toys?

Improvementsin medical imaging and a steady increase in computing power are leading to new possibilities in the field of cardiovascular interventions. Interventions can be planned in advance in greater detail, even to the point of simulating procedures. Nevertheless, all techniques are at an early stage of development. It is of utmost importance that tools, especially if they can be used as decision support are intensively validated and their accuracy is demonstrated. In our commentary, we summarize current techniques for impprovements in planning and guiding of procedures, but also critically discuss the downsides of these techniques. Following the work of Kenichi and colleagues, we also discuss necessary steps in advancing new tools and techniques, particularly as they are used in routine clinical practice. We also discuss the role of artificial intelligence, which could play a crucial role in this context in the future.

Deep learning methods for automatic evaluation of delayed enhancement-MRI. The results of the EMIDEC challenge.

A key factor for assessing the state of the heart after myocardial infarction (MI) is to measure whether the myocardium segment is viable after reperfusion or revascularization therapy. Delayed enhancement-MRI or DE-MRI, which is performed 10 min after injection of the contrast agent, provides high contrast between viable and nonviable myocardium and is therefore a method of choice to evaluate the extent of MI. To automatically assess myocardial status, the results of the EMIDEC challenge that focused on this task are presented in this paper. The challenge's main objectives were twofold. First, to evaluate if deep learning methods can distinguish between non-infarct and pathological exams, i.e. exams with or without hyperenhanced area. Second, to automatically calculate the extent of myocardial infarction. The publicly available database consists of 150 exams divided into 50 cases without any hyperenhanced area after injection of a contrast agent and 100 cases with myocardial infarction (and then with a hyperenhanced area on DE-MRI), whatever their inclusion in the cardiac emergency department. Along with MRI, clinical characteristics are also provided. The obtained results issued from several works show that the automatic classification of an exam is a reachable task (the best method providing an accuracy of 0.92), and the automatic segmentation of the myocardium is possible. However, the segmentation of the diseased area needs to be improved, mainly due to the small size of these areas and the lack of contrast with the surrounding structures.

Toward Sharing Brain Images: Differentially Private TOF-MRA Images With Segmentation Labels Using Generative Adversarial Networks.

Sharing labeled data is crucial to acquire large datasets for various Deep Learning applications. In medical imaging, this is often not feasible due to privacy regulations. Whereas anonymization would be a solution, standard techniques have been shown to be partially reversible. Here, synthetic data using a Generative Adversarial Network (GAN) with differential privacy guarantees could be a solution to ensure the patient's privacy while maintaining the predictive properties of the data. In this study, we implemented a Wasserstein GAN (WGAN) with and without differential privacy guarantees to generate privacy-preserving labeled Time-of-Flight Magnetic Resonance Angiography (TOF-MRA) image patches for brain vessel segmentation. The synthesized image-label pairs were used to train a U-net which was evaluated in terms of the segmentation performance on real patient images from two different datasets. Additionally, the Fréchet Inception Distance (FID) was calculated between the generated images and the real images to assess their similarity. During the evaluation using the U-Net and the FID, we explored the effect of different levels of privacy which was represented by the parameter ϵ. With stricter privacy guarantees, the segmentation performance and the similarity to the real patient images in terms of FID decreased. Our best segmentation model, trained on synthetic and private data, achieved a Dice Similarity Coefficient (DSC) of 0.75 for ϵ = 7.4 compared to 0.84 for ϵ = ∞ in a brain vessel segmentation paradigm (DSC of 0.69 and 0.88 on the second test set, respectively). We identified a threshold of ϵ <5 for which the performance (DSC <0.61) became unstable and not usable. Our synthesized labeled TOF-MRA images with strict privacy guarantees retained predictive properties necessary for segmenting the brain vessels. Although further research is warranted regarding generalizability to other imaging modalities and performance improvement, our results mark an encouraging first step for privacy-preserving data sharing in medical imaging.

A Collaborative Approach for the Development and Application of Machine Learning Solutions for CMR-Based Cardiac Disease Classification.

The quality and acceptance of machine learning (ML) approaches in cardiovascular data interpretation depends strongly on model design and training and the interaction with the clinical experts. We hypothesize that a software infrastructure for the training and application of ML models can support the improvement of the model training and provide relevant information for understanding the classification-relevant data features. The presented solution supports an iterative training, evaluation, and exploration of machine-learning-based multimodal data interpretation methods considering cardiac MRI data. Correction, annotation, and exploration of clinical data and interpretation of results are supported through dedicated interactive visual analytics tools. We test the presented concept with two use cases from the ACDC and EMIDEC cardiac MRI image analysis challenges. In both applications, pre-trained 2D U-Nets are used for segmentation, and classifiers are trained for diagnostic tasks using radiomics features of the segmented anatomical structures. The solution was successfully used to identify outliers in automatic segmentation and image acquisition. The targeted curation and addition of expert annotations improved the performance of the machine learning models. Clinical experts were supported in understanding specific anatomical and functional characteristics of the assigned disease classes.

Generating 3D TOF-MRA volumes and segmentation labels using generative adversarial networks.

Deep learning requires large labeled datasets that are difficult to gather in medical imaging due to data privacy issues and time-consuming manual labeling. Generative Adversarial Networks (GANs) can alleviate these challenges enabling synthesis of shareable data. While 2D GANs have been used to generate 2D images with their corresponding labels, they cannot capture the volumetric information of 3D medical imaging. 3D GANs are more suitable for this and have been used to generate 3D volumes but not their corresponding labels. One reason might be that synthesizing 3D volumes is challenging owing to computational limitations. In this work, we present 3D GANs for the generation of 3D medical image volumes with corresponding labels applying mixed precision to alleviate computational constraints. We generated 3D Time-of-Flight Magnetic Resonance Angiography (TOF-MRA) patches with their corresponding brain blood vessel segmentation labels. We used four variants of 3D Wasserstein GAN (WGAN) with: 1) gradient penalty (GP), 2) GP with spectral normalization (SN), 3) SN with mixed precision (SN-MP), and 4) SN-MP with double filters per layer (c-SN-MP). The generated patches were quantitatively evaluated using the Fréchet Inception Distance (FID) and Precision and Recall of Distributions (PRD). Further, 3D U-Nets were trained with patch-label pairs from different WGAN models and their performance was compared to the performance of a benchmark U-Net trained on real data. The segmentation performance of all U-Net models was assessed using Dice Similarity Coefficient (DSC) and balanced Average Hausdorff Distance (bAVD) for a) all vessels, and b) intracranial vessels only. Our results show that patches generated with WGAN models using mixed precision (SN-MP and c-SN-MP) yielded the lowest FID scores and the best PRD curves. Among the 3D U-Nets trained with synthetic patch-label pairs, c-SN-MP pairs achieved the highest DSC (0.841) and lowest bAVD (0.508) compared to the benchmark U-Net trained on real data (DSC 0.901; bAVD 0.294) for intracranial vessels. In conclusion, our solution generates realistic 3D TOF-MRA patches and labels for brain vessel segmentation. We demonstrate the benefit of using mixed precision for computational efficiency resulting in the best-performing GAN-architecture. Our work paves the way towards sharing of labeled 3D medical data which would increase generalizability of deep learning models for clinical use.

Fully automated quantification of in vivo viscoelasticity of prostate zones using magnetic resonance elastography with Dense U-net segmentation.

Magnetic resonance elastography (MRE) for measuring viscoelasticity heavily depends on proper tissue segmentation, especially in heterogeneous organs such as the prostate. Using trained network-based image segmentation, we investigated if MRE data suffice to extract anatomical and viscoelastic information for automatic tabulation of zonal mechanical properties of the prostate. Overall, 40 patients with benign prostatic hyperplasia (BPH) or prostate cancer (PCa) were examined with three magnetic resonance imaging (MRI) sequences: T2-weighted MRI (T2w), diffusion-weighted imaging (DWI), and MRE-based tomoelastography, yielding six independent sets of imaging data per patient (T2w, DWI, apparent diffusion coefficient, MRE magnitude, shear wave speed, and loss angle maps). Combinations of these data were used to train Dense U-nets with manually segmented masks of the entire prostate gland (PG), central zone (CZ), and peripheral zone (PZ) in 30 patients and to validate them in 10 patients. Dice score (DS), sensitivity, specificity, and Hausdorff distance were determined. We found that segmentation based on MRE magnitude maps alone (DS, PG: 0.93 ± 0.04, CZ: 0.95 ± 0.03, PZ: 0.77 ± 0.05) was more accurate than magnitude maps combined with T2w and DWI_b (DS, PG: 0.91 ± 0.04, CZ: 0.91 ± 0.06, PZ: 0.63 ± 0.16) or T2w alone (DS, PG: 0.92 ± 0.03, CZ: 0.91 ± 0.04, PZ: 0.65 ± 0.08). Automatically tabulated MRE values were not different from ground-truth values (P>0.05). In conclusion, MRE combined with Dense U-net segmentation allows tabulation of quantitative imaging markers without manual analysis and independent of other MRI sequences and can thus contribute to PCa detection and classification.

Detection and analysis of cerebral aneurysms based on X-ray rotational angiography - the CADA 2020 challenge.

The Cerebral Aneurysm Detection and Analysis (CADA) challenge was organized to support the development and benchmarking of algorithms for detecting, analyzing, and risk assessment of cerebral aneurysms in X-ray rotational angiography (3DRA) images. 109 anonymized 3DRA datasets were provided for training, and 22 additional datasets were used to test the algorithmic solutions. Cerebral aneurysm detection was assessed using the F2 score based on recall and precision, and the fit of the delivered bounding box was assessed using the distance to the aneurysm. The segmentation quality was measured using the Jaccard index and a combination of different surface distance measures. Systematic errors were analyzed using volume correlation and bias. Rupture risk assessment was evaluated using the F2 score. 158 participants from 22 countries registered for the CADA challenge. The U-Net-based detection solutions presented by the community show similar accuracy compared to experts (F2 score 0.92), with a small number of missed aneurysms with diameters smaller than 3.5 mm. In addition, the delineation of these structures, based on U-Net variations, is excellent, with a Jaccard score of 0.92. The rupture risk estimation methods achieved an F2 score of 0.71. The performance of the detection and segmentation solutions is equivalent to that of human experts. The best results are obtained in rupture risk estimation by combining different image-based, morphological, and computational fluid dynamic parameters using machine learning methods. Furthermore, we evaluated the best methods pipeline, from detecting and delineating the vessel dilations to estimating the risk of rupture. The chain of these methods achieves an F2-score of 0.70, which is comparable to applying the risk prediction to the ground-truth delineation (0.71).

Cascaded neural network-based CT image processing for aortic root analysis.

PURPOSE: Careful assessment of the aortic root is paramount to select an appropriate prosthesis for transcatheter aortic valve implantation (TAVI). Relevant information about the aortic root anatomy, such as the aortic annulus diameter, can be extracted from pre-interventional CT. In this work, we investigate a neural network-based approach for segmenting the aortic root as a basis for obtaining these parameters. METHODS: To support valve prosthesis selection, geometric measures of the aortic root are extracted from the patient's CT scan using a cascade of convolutional neural networks (CNNs). First, the image is reduced to the aortic root, valve, and left ventricular outflow tract (LVOT); within that subimage, the aortic valve and ascending aorta are segmented; and finally, the region around the aortic annulus. From the segmented annulus region, we infer the annulus orientation using principal component analysis (PCA). The area-derived diameter of the annulus is approximated based on the segmentation of the aortic root and LVOT and the plane orientation resulting from the PCA. RESULTS: The cascade of CNNs was trained using 90 expert-annotated contrast-enhanced CT scans routinely acquired for TAVI planning. Segmentation of the aorta and valve within the region of interest achieved an F1 score of 0.94 on the test set of 36 patients. The area-derived diameter within the annulus region was determined with a mean error below 2 mm between the automatic measurement and the diameter derived from annotations. The calculated diameters and resulting errors are comparable to published results of alternative approaches. CONCLUSIONS: The cascaded neural network approach enabled the assessment of the aortic root with a relatively small training set. The processing time amounts to 30 s per patient, facilitating time-efficient, reproducible measurements. An extended training data set, including different levels of calcification or special cases (e.g., pre-implanted valves), could further improve this method's applicability and robustness.