Hybrid 3D printing: a game-changer in personalized cardiac medicine?

Three-dimensional (3D) printing in congenital heart disease has the potential to increase procedural efficiency and patient safety by improving interventional and surgical planning and reducing radiation exposure. Cardiac magnetic resonance imaging and computed tomography are usually the source datasets to derive 3D printing. More recently, 3D echocardiography has been demonstrated to derive 3D-printed models. The integration of multiple imaging modalities for hybrid 3D printing has also been shown to create accurate printed heart models, which may prove to be beneficial for interventional cardiologists, cardiothoracic surgeons, and as an educational tool. Further advancements in the integration of different imaging modalities into a single platform for hybrid 3D printing and virtual 3D models will drive the future of personalized cardiac medicine.     

基于超声的3D打印技术在心脏领域的应用进展

3D打印技术是近年来出现的新技术。随着三维超声的发展,基于超声的3D打印技术在心血管疾病诊断和治疗中逐渐被应用。3D打印技术通过获取图像、建模、实体打印三个步骤可以把三维超声图像转换为实体模型。目前基于超声的3D打印技术主要应用在左心耳封堵术、瓣膜性心脏病、先天性心脏病三个方面,有助于术前评估、术前模拟手术、医疗装置设计、血流动力学模拟及医学沟通教育。3D打印技术应用前景广阔,打印具有生物活性的组织或者结构直接应用于人体是未来的发展方向。为了让3D打印技术更好地服务临床,我们仍面临着巨大的挑战。     

3D打印技术辅助诊断和指导室间隔缺损手术

目的 探讨运用3D打印技术构建三维心脏模型对结构性心脏病诊断及治疗的可行性与临床价值。方法 基于1例室间隔缺损患儿的心脏CT影像数据,采用Mimics 17.0图形分割软件,以阈值法进行心肌分割,以动态区域生长法进行血池分割获得数字化三维心脏模型;采用PolyJet材料,以Objet500 Connex3聚合物喷射成型3D打印机打印软质的实体心脏模型,并对心脏模型进行评估。结果 3D模型可清晰显示心脏内部解剖结构、空间关系以及室间隔缺损的三维形态。在打印出的实体心脏模型上测量室间隔缺损的轴位长度为4.85 mm,与3D数字模型和CT图像中测量结果（轴位长度分别为4.86 mm和4.84 mm,矢状位长度分别为6.66 mm和6.65 mm）基本一致。结论 将3D打印技术用于结构性心脏病有助于诊断和制定适当的手术方案;通过规范化的图像处理技术获取精确的心脏模型,使个性化精准诊疗结构性心脏病成为可能。     

Multi-institutional evaluation of the indications and radiation dose of functional cardiovascular computed tomography (CCT) imaging in congenital heart disease

Ventricular volumes and ejection fraction are often used in clinical decision making in patients with congenital heart disease (CHD). The referral diagnosis, radiation exposure and image quality of functional cardiac computed tomography (CT) in a relatively large cohort of patients of CHD has not been reported. This is a retrospective evaluation of functional CT studies performed in CHD patients from three institutions (1/2007–3/2013). Patient and scanner characteristics, radiation dose estimates and image quality were compared. Two hundred ninety-eight functional CT studies were evaluated. The most common referral diagnosis were tetralogy of Fallot (33 %), transposition complexes (24 %) single ventricle heart disease (15 %), and left sided obstruction (15 %). The reason for cardiac CT was presence of pacemaker (60 %), need for detailed coronary artery imaging (18 %), metallic artifact in CMR (12 %), evaluation of prosthetic valve function (4 %), and claustrophobia or BMI too large for the available MR scanner (6 %). 266 (89.3 %) scans allowed quantification of ventricular function, 25 (8.4 %) scans allowed qualitative assessment of function, and 7 (2.3 %) of the scans were non-diagnostic for functional analysis. Median DLP was 399 mGy cm (186, 614), and median effective dose was 5.5 mSv (2.6, 8.5). Radiation dose and image quality varied across institutions. Cardiac CT function imaging can be performed in patients with congenital heart disease when CMR is contraindicated or has poor image quality. Radiation dose and image quality varies across institutions.     

基于经食管三维超声的3D打印模型评价左心耳解剖

目的 探讨经食管三维超声心动图（3D-TEE）作为左心耳（LAA）三维（3D）打印数据源的可行性及超声3D打印模型的准确性,并评价3D打印模型对特殊解剖形态LAA封堵手术的指导价值。方法 回顾性分析18例因心房颤动接受LAA封堵术的患者资料,包括LAA的3D-TEE和CT容积图像数据。对3D-TEE数据进行后处理,并使用弹性材料制作LAA的3D打印模型。采用3D打印模型评价LAA解剖分型与分叶分型,测量LAA开口的最大径、最小径及深度,并与3D-TEE及CT容积图像结果进行对比。在封堵困难的特殊形态LAA模型上进行封堵器释放试验。结果 对18例患者均成功进行超声数据后处理并获取LAA 3D打印模型。应用3D打印模型与CT容积图像对LAA进行解剖分型及分叶分型的一致性均较高,Kappa值分别为0.92和0.83。且3D打印模型对LAA开口最大径、最小径及深度的测量值与3D-TEE测量值差异均无统计学意义（P均>0.05）。18例中2例为特殊形态LAA,均通过3D打印模型进行封堵伞释放模拟出术中封堵过程。结论 基于3D-TEE的LAA 3D打印技术具有较高的可行性及准确性,有助于指导特殊形态LAA的个性化封堵。     

CT based 3D printing is superior to transesophageal echocardiography for pre-procedure planning in left atrial appendage device closure

Accurate assessment of the left atrial appendage (LAA) is important for pre-procedure planning when utilizing device closure for stroke reduction. Sizing is traditionally done with transesophageal echocardiography (TEE) but this is not always precise. Three-dimensional (3D) printing of the LAA may be more accurate. 24 patients underwent Watchman device (WD) implantation (71?±?11 years, 42% female). All had complete 2-dimensional TEE. Fourteen also had cardiac computed tomography (CCT) with 3D printing to produce a latex model of the LAA for pre-procedure planning. Device implantation was unsuccessful in 2 cases (one with and one without a 3D model). The model correlated perfectly with implanted device size (R²?=?1; p?<?0.001), while TEE-predicted size showed inferior correlation (R²?=?0.34; 95% CI 0.23–0.98, p?=?0.03). Fisher’s exact test showed the model better predicted final WD size than TEE (100 vs. 60%, p?=?0.02). Use of the model was associated with reduced procedure time (70?±?20 vs. 107?±?53 min, p?=?0.03), anesthesia time (134?±?31 vs. 182?±?61 min, p?=?0.03), and fluoroscopy time (11?±?4 vs. 20?±?13 min, p?=?0.02). Absence of peri-device leak was also more likely when the model was used (92 vs. 56%, p?=?0.04). There were trends towards reduced trans-septal puncture to catheter removal time (50?±?20 vs. 73?±?36 min, p?=?0.07), number of device deployments (1.3?±?0.5 vs. 2.0?±?1.2, p?=?0.08), and number of devices used (1.3?±?0.5 vs. 1.9?±?0.9, p?=?0.07). Patient specific models of the LAA improve precision in closure device sizing. Use of the printed model allowed rapid and intuitive location of the best landing zone for the device.     

4D modelling for rapid assessment of biventricular function in congenital heart disease

Although more patients with congenital heart disease (CHD) are now living longer due to better surgical interventions, they require regular imaging to monitor cardiac performance. There is a need for robust clinical tools which can accurately assess cardiac function of both the left and right ventricles in these patients. We have developed methods to rapidly quantify 4D (3D + time) biventricular function from standard cardiac MRI examinations. A finite element model was interactively customized to patient images using guide-point modelling. Computational efficiency and ability to model large deformations was improved by predicting cardiac motion for the left ventricle and epicardium with a polar model. In addition, large deformations through the cycle were more accurately modeled using a Cartesian deformation penalty term. The model was fitted to user-defined guide points and image feature tracking displacements throughout the cardiac cycle. We tested the methods in 60 cases comprising a variety of congenital heart diseases and showed good correlation with the gold standard manual analysis, with acceptable inter-observer error. The algorithm was considerably faster than standard analysis and shows promise as a clinical tool for patients with CHD.     

The production of digital and printed resources from multiple modalities using visualization and three-dimensional printing techniques

Purpose

Virtual digital resources and printed models have become indispensable tools for medical training and surgical planning. Nevertheless, printed models of soft tissue organs are still challenging to reproduce. This study adopts open source packages and a low-cost desktop 3D printer to convert multiple modalities of medical images to digital resources (volume rendering images and digital models) and lifelike printed models, which are useful to enhance our understanding of the geometric structure and complex spatial nature of anatomical organs.

Materials and methods

Neuroimaging technologies such as CT, CTA, MRI, and TOF-MRA collect serial medical images. The procedures for producing digital resources can be divided into volume rendering and medical image reconstruction. To verify the accuracy of reconstruction, this study presents qualitative and quantitative assessments. Subsequently, digital models are archived as stereolithography format files and imported to the bundled software of the 3D printer. The printed models are produced using polylactide filament materials.

Results

We have successfully converted multiple modalities of medical images to digital resources and printed models for both hard organs (cranial base and tooth) and soft tissue organs (brain, blood vessels of the brain, the heart chambers and vessel lumen, and pituitary tumor). Multiple digital resources and printed models were provided to illustrate the anatomical relationship between organs and complicated surrounding structures. Three-dimensional printing (3DP) is a powerful tool to produce lifelike and tangible models.

Conclusions

We present an available and cost-effective method for producing both digital resources and printed models. The choice of modality in medical images and the processing approach is important when reproducing soft tissue organs models. The accuracy of the printed model is determined by the quality of organ models and 3DP. With the ongoing improvement of printing techniques and the variety of materials available, 3DP will become an indispensable tool in medical training and surgical planning.

     

3D Printing from Cardiac Computed Tomography for Procedural Planning

Purpose of Review

Computed tomography (CT)-based three-dimensional (3D) printing is an emerging field in preoperative visualization of the cardiac anatomy, procedure planning, and simulations of heart surgeries. This article summarizes the uses of 3D-printed models in preoperative cardiac procedure planning and intraprocedural navigation during surgeries organized on a case-by-case basis.

Recent Findings

The 3D models have recently been shown to aid in a variety of procedures. Some of these include valvular replacement or repair, management of paravalvular leaks, left atrial appendage closure, myectomy in hypertrophic cardiomyopathy, cardiac aneurysm repair, tumor resection, and surgical planning in congenital heart lesions.

Summary

We anticipate that 3D models will become more accessible, and their utilization will increase in the near future. However, CT-based 3D printing will require further assessment with multicenter-based studies. Multicenter explorations will allow for consensus and standardization of such potentially useful tools, given the vast availability of printing and imaging techniques currently available.

     

Ventricular pressure–volume loops obtained by 3D real-time echocardiography and mini pressure wire—a feasibility study

Introduction

Pressure–volume relations (PVR) provide vital information regarding ventricular performance and cardiac pathophysiology. Acquiring PVR by conductance catheter technology is invasive and laborious, which explains why the assessment of PVR is not used in clinical practice. Real-time three-dimensional echocardiography (3DE) allows almost instantaneous capture of ventricular volume changes throughout the cardiac cycle. The aim of the study was to assess the feasibility of 3DE combined with pressure data to calculate PVR in children and adolescents.

Methods

In 31 patients with congenital heart disease (age 3 days–22.7 years, weight 2.75–80.0 kg), ventricular pressure was recorded by a mini pressure wire during routine catheterization. Simultaneously, 3D datasets of the left or right ventricle were acquired for calculation of volume. PVR were generated from contemporaneous 3D volume and pressure data. Systolic myocardial elastance, ventriculo-arterial coupling, diastolic relaxation constant Tau and end-diastolic PVR were determined using a single-beat approach.

Results

Computation of PVR using non-invasive 3D volume data and pressure curves obtained by mini pressure wire was easy, feasible and reproducible. On average, 6 [3–11] PVR, needing an additional examination time of 6.5 ± 3.5 min, were acquired. Both intra- and interobserver variability were good for all measured parameters (coefficient of variation <10%).

Conclusions

Calculation of PVR from 3DE volume curves and simultaneous pressure data obtained by a mini pressure wire is a feasible method to assess cardiac function. Due to the tiny size of the pressure wire used, PVR can be acquired even in small neonates with congenital heart disease.     

Learned iterative segmentation of highly variable anatomy from limited data: Applications to whole heart segmentation for congenital heart disease

Training deep learning models that segment an image in one step typically requires a large collection of manually annotated images that captures the anatomical variability in a cohort. This poses challenges when anatomical variability is extreme but training data is limited, as when segmenting cardiac structures in patients with congenital heart disease (CHD). In this paper, we propose an iterative segmentation model and show that it can be accurately learned from a small dataset. Implemented as a recurrent neural network, the model evolves a segmentation over multiple steps, from a single user click until reaching an automatically determined stopping point. We develop a novel loss function that evaluates the entire sequence of output segmentations, and use it to learn model parameters. Segmentations evolve predictably according to growth dynamics encapsulated by training data, which consists of images, partially completed segmentations, and the recommended next step. The user can easily refine the final segmentation by examining those that are earlier or later in the output sequence. Using a dataset of 3D cardiac MR scans from patients with a wide range of CHD types, we show that our iterative model offers better generalization to patients with the most severe heart malformations.     

Direct 3‐D printing of Ti‐6Al‐4V/HA composite porous scaffolds for customized mechanical properties and biological functions

Customized scaffold plays an important role in bone tissue regeneration. Precise control of the mechanical properties and biological functions of scaffolds still remains a challenge. In this study, metal and ceramic biomaterials are composited by direct 3‐D printing. Hydroxyapatite (HA) powders with diameter of about 25 μm and Ti‐6Al‐4V powders with diameter of 15–53 μm were mixed and modulated for preparing 3‐D printing inks formulation. Three different proportions of 8, 10, and 25 wt.% HA specimens were printed with same porosity of 72.1%. The green bodies of the printed porous scaffolds were sintered at 1,150°C in the atmosphere of argon furnace and conventional muffle furnace. The porosities of the final 3‐D‐printed specimens were 64.3 ± 0.8% after linear shrinkage of 6.5 ± 0.8%. The maximum compressive strength of the 3‐D‐printed scaffolds can be flexibly customized in a wide range. The maximum compressive strength of these scaffolds in this study ranged from 3.07 to 60.4 MPa, depending on their different preparation process. The phase composition analysis and microstructure characterization indicated that the Ti‐6Al‐4V and HA were uniformly composited in the scaffolds. The cytocompatibility and osteogenic properties were evaluated in vitro with rabbit bone marrow stromal cells (rBMSCs). Differentiation and proliferation of rBMSCs indicated good biocompatibility of the 3‐D‐printed scaffolds. The proposed 3‐D printing of Ti‐6Al‐4V/HA composite porous scaffolds with tunable mechanical and biological properties in this study is a promising candidate for bone tissue engineering.     

Semiautomatic classification of acetabular shape from three-dimensional ultrasound for diagnosis of infant hip dysplasia using geometric features

Purpose

Developmental dysplasia of the hip (DDH) is a congenital deformity which in severe cases leads to hip dislocation and in milder cases to premature osteoarthritis. Image-aided diagnosis of DDH is partly based on Graf classification which quantifies the acetabular shape seen at two-dimensional ultrasound (2DUS), which leads to high inter-scan variance. 3D ultrasound (3DUS) is a promising alternative for more reliable DDH diagnosis. However, manual quantification of acetabular shape from 3DUS is cumbersome.

Methods

Here, we (1) propose a semiautomated segmentation algorithm to delineate 3D acetabular surface models from 3DUS using graph search; (2) propose a fully automated method to classify acetabular shape based on a random forest (RF) classifier using features derived from 3D acetabular surface models; and (3) test diagnostic accuracy on a dataset of 79 3DUS infant hip recordings (36 normal, 16 borderline, 27 dysplastic based on orthopedic surgeon assessment) in 42 patients. For each 3DUS, we performed semiautomated segmentation to produce 3D acetabular surface models and then calculated geometric features including the automatic \(\mathrm{a}\)lpha (AA), acetabular contact angle (ACA), kurtosis (K), skewness (S) and convexity (C). Mean values of features obtained from surface models were used as inputs to train a RF classifier.

Results

Surface models were generated rapidly (user time 46.2 s) via semiautomated segmentation and visually closely correlated with actual acetabular contours (RMS error 1.39 ± 0.7 mm). A paired nonparametric u test on of feature values in each category showed statistically significant variation (p < 0.001) for AA, ACA and convexity. The RF classifier was 100 % specific and 97.2 % sensitive in classifying normal versus dysplastic hips and yielded true positive rates of 94.4, 62.5 and 89.9 % for normal, borderline and dysplastic hips.

Conclusions

The proposed technique reduces the subjectivity of image-aided DDH diagnosis and could be useful in clinical practice.

     

Semiautomatic three-dimensional CT ventricular volumetry in patients with congenital heart disease: agreement between two methods with different user interaction

To assess agreement between two semi-automatic, three-dimensional (3D) computed tomography (CT) ventricular volumetry methods with different user interactions in patients with congenital heart disease. In 30 patients with congenital heart disease (median age 8 years, range 5 days–33 years; 20 men), dual-source, multi-section, electrocardiography-synchronized cardiac CT was obtained at the end-systolic (n = 22) and/or end-diastolic (n = 28) phase. Nineteen left ventricle end-systolic (LV ESV), 28 left ventricle end-diastolic (LV EDV), 22 right ventricle end-systolic (RV ESV), and 28 right ventricle end-diastolic volumes (RV EDV) were successfully calculated using two semi-automatic, 3D segmentation methods with different user interactions (high in method 1, low in method 2). The calculated ventricular volumes of the two methods were compared and correlated. A P value <0.05 was considered statistically significant. LV ESV (35.95 ± 23.49 ml), LV EDV (88.76 ± 61.83 ml), and RV ESV (46.87 ± 47.39 ml) measured by method 2 were slightly but significantly smaller than those measured by method 1 (41.25 ± 26.94 ml, 92.20 ± 62.69 ml, 53.61 ± 50.08 ml for LV ESV, LV EDV, and RV ESV, respectively; P ≤ 0.02). In contrast, no statistically significant difference in RV EDV (122.57 ± 88.57 ml in method 1, 123.83 ± 89.89 ml in method 2; P = 0.36) was found between the two methods. All ventricular volumes showed very high correlation (R = 0.978, 0.993, 0.985, 0.997 for LV ESV, LV EDV, RV ESV, and RV EDV, respectively; P < 0.001) between the two methods. In patients with congenital heart disease, 3D CT ventricular volumetry shows good agreement and high correlation between the two methods, but method 2 tends to slightly underestimate LV ESV, LV EDV, and RV ESV.     

Real-time three dimensional CT and MRI to guide interventions for congenital heart disease and acquired pulmonary vein stenosis

To validate the feasibility and spatial accuracy of pre-procedural 3D images to 3D rotational fluoroscopy registration to guide interventional procedures in patients with congenital heart disease and acquired pulmonary vein stenosis. Cardiac interventions in patients with congenital and structural heart disease require complex catheter manipulation. Current technology allows registration of the anatomy obtained from 3D CT and/or MRI to be overlaid onto fluoroscopy. Thirty patients scheduled for interventional procedures from 12/2012 to 8/2015 were prospectively recruited. A C-arm CT using a biplane C-arm system (Artis zee, VC14H, Siemens Healthcare) was acquired to enable 3D3D registration with pre-procedural images. Following successful image fusion, the anatomic landmarks marked in pre-procedural images were overlaid on live fluoroscopy. The accuracy of image registration was determined by measuring the distance between overlay markers and a reference point in the image. The clinical utility of the registration was evaluated as either “High”, “Medium” or “None”. Seventeen patients with congenital heart disease and 13 with acquired pulmonary vein stenosis were enrolled. Accuracy and benefit of registration were not evaluated in two patients due to suboptimal images. The distance between the marker and the actual anatomical location was 0–2 mm in 18 (64%), 2–4 mm in 3 (11%) and >4 mm in 7 (25%) patients. 3D3D registration was highly beneficial in 18 (64%), intermediate in 3 (11%), and not beneficial in 7 (25%) patients. 3D3D registration can facilitate complex congenital and structural interventions. It may reduce procedure time, radiation and contrast dose.     

A Fully-Automatic Method to Segment the Carotid Artery Layers in Ultrasound Imaging: Application to Quantify the Compression-Decompression Pattern of the Intima-Media Complex During the Cardiac Cycle

The aim of this study was to introduce and evaluate a contour segmentation method to extract the interfaces of the intima–media complex in carotid B-mode ultrasound images. The method was applied to assess the temporal variation of intima–media thickness during the cardiac cycle. The main methodological contribution of the proposed approach is the introduction of an augmented dimension to process 2-D images in a 3-D space. The third dimension, which is added to the two spatial dimensions of the image, corresponds to the tentative local thickness of the intima–media complex. The method is based on a dynamic programming scheme that runs in a 3-D space generated with a shape-adapted filter bank. The optimal solution corresponds to a single medial axis representation that fully describes the two anatomical interfaces of the arterial wall. The method is fully automatic and does not require any input from the user. The method was trained on 60 subjects and validated on 184 other subjects from six different cohorts and four different medical centers. The arterial wall was successfully segmented in all analyzed images (average pixel size = 57 ± 20 μm), with average segmentation errors of 47 ± 70 μm for the lumen–intima interface, 55 ± 68 μm for the media–adventitia interface and 66 ± 90 μm for the intima–media thickness. The amplitude of the temporal variations in IMT during the cardiac cycle was significantly higher in the diseased population than in healthy volunteers (106 ± 48 vs. 86 ± 34 μm, p = 0.001). The introduced framework is a promising approach to investigate an emerging functional parameter of the arterial wall by assessing the cyclic compression–decompression pattern of the tissues.     

3D X-ray imaging methods in support catheter ablations of cardiac arrhythmias

Cardiac arrhythmias are a very frequent illness. Pharmacotherapy is not very effective in persistent arrhythmias and brings along a number of risks. Catheter ablation has became an effective and curative treatment method over the past 20 years. To support complex arrhythmia ablations, the 3D X-ray cardiac cavities imaging is used, most frequently the 3D reconstruction of CT images. The 3D cardiac rotational angiography (3DRA) represents a modern method enabling to create CT like 3D images on a standard X-ray machine equipped with special software. Its advantage lies in the possibility to obtain images during the procedure, decreased radiation dose and reduction of amount of the contrast agent. The left atrium model is the one most frequently used for complex atrial arrhythmia ablations, particularly for atrial fibrillation. CT data allow for creation and segmentation of 3D models of all cardiac cavities. Recently, a research has been made proving the use of 3DRA to create 3D models of other cardiac (right ventricle, left ventricle, aorta) and non-cardiac structures (oesophagus). They can be used during catheter ablation of complex arrhythmias to improve orientation during the construction of 3D electroanatomic maps, directly fused with 3D electroanatomic systems and/or fused with fluoroscopy. An intensive development in the 3D model creation and use has taken place over the past years and they became routinely used during catheter ablations of arrhythmias, mainly atrial fibrillation ablation procedures. Further development may be anticipated in the future in both the creation and use of these models.     

Subject‐specific aortic wall shear stress estimations using semi‐automatic segmentation

Atherosclerosis development is strongly believed to be influenced by hemodynamic forces such as wall shear stress (WSS). To estimate such an entity in‐vivo in humans, image‐based computational fluid dynamics (CFD) is a useful tool. In this study, we use a combination of magnetic resonance imaging (MRI) and CFD to estimate WSS. In such method, a number of steps are included. One important step is the interpretation of images into 3D models, named segmentation. The choice of segmentation method can influence the resulting WSS distribution in the human aorta. This is studied by comparing WSS results gained from the use of two different segmentation approaches: manual and semi‐automatic, where the manual approach is considered to be the reference method. The investigation is performed on a group of eight healthy male volunteers. The different segmentation methods give slightly different geometrical depictions of the human aorta (difference in the mean thoracic Aorta lumen diameter were 0·7% P<0·86). However, there is a very good agreement between the resulting WSS distribution for the two segmentation approaches. The small differences in WSS between the methods increase in the late systole and early diastolic cardiac cycle time point indicating that the WSS is more sensitive to local geometric differences in these parts of the cardiac cycle (correlation coefficient is 0·96 at peak systole and 0·68 at early diastole). We can conclude that the results show that the semi‐automatic segmentation method can be used in future to estimate relevant aortic WSS.     

Motion-aware stroke volume quantification in 4D PC-MRI data of the human aorta

Purpose

4D PC-MRI enables the noninvasive measurement of time-resolved, three-dimensional blood flow data that allow quantification of the hemodynamics. Stroke volumes are essential to assess the cardiac function and evolution of different cardiovascular diseases. The calculation depends on the wall position and vessel orientation, which both change during the cardiac cycle due to the heart muscle contraction and the pumped blood. However, current systems for the quantitative 4D PC-MRI data analysis neglect the dynamic character and instead employ a static 3D vessel approximation. We quantify differences between stroke volumes in the aorta obtained with and without consideration of its dynamics.

Methods

We describe a method that uses the approximating 3D segmentation to automatically initialize segmentation algorithms that require regions inside and outside the vessel for each temporal position. This enables the use of graph cuts to obtain 4D segmentations, extract vessel surfaces including centerlines for each temporal position and derive motion information. The stroke volume quantification is compared using measuring planes in static (3D) vessels, planes with fixed angulation inside dynamic vessels (this corresponds to the common 2D PC-MRI) and moving planes inside dynamic vessels.

Results

Seven datasets with different pathologies such as aneurysms and coarctations were evaluated in close collaboration with radiologists. Compared to the experts’ manual stroke volume estimations, motion-aware quantification performs, on average, 1.57 % better than calculations without motion consideration. The mean difference between stroke volumes obtained with the different methods is 7.82 %. Automatically obtained 4D segmentations overlap by 85.75 % with manually generated ones.

Conclusion

Incorporating motion information in the stroke volume quantification yields slight but not statistically significant improvements. The presented method is feasible for the clinical routine, since computation times are low and essential parts run fully automatically. The 4D segmentations can be used for other algorithms as well. The simultaneous visualization and quantification may support the understanding and interpretation of cardiac blood flow.