Comprehensive Modeling and Visualization of Cardiac Anatomy and Physiology from CT Imaging and Computer Simulations.

In clinical cardiology, both anatomy and physiology are needed to diagnose cardiac pathologies. CT imaging and computer simulations provide valuable and complementary data for this purpose. However, it remains challenging to gain useful information from the large amount of high-dimensional diverse data. The current tools are not adequately integrated to visualize anatomic and physiologic data from a complete yet focused perspective. We introduce a new computer-aided diagnosis framework, which allows for comprehensive modeling and visualization of cardiac anatomy and physiology from CT imaging data and computer simulations, with a primary focus on ischemic heart disease. The following visual information is presented: (1) Anatomy from CT imaging: geometric modeling and visualization of cardiac anatomy, including four heart chambers, left and right ventricular outflow tracts, and coronary arteries; (2) Function from CT imaging: motion modeling, strain calculation, and visualization of four heart chambers; (3) Physiology from CT imaging: quantification and visualization of myocardial perfusion and contextual integration with coronary artery anatomy; (4) Physiology from computer simulation: computation and visualization of hemodynamics (e.g., coronary blood velocity, pressure, shear stress, and fluid forces on the vessel wall). Substantially, feedback from cardiologists have confirmed the practical utility of integrating these features for the purpose of computer-aided diagnosis of ischemic heart disease.

Watertight modeling and segmentation of bifurcated Coronary arteries for blood flow simulation using CT imaging.

Image-based simulation of blood flow using computational fluid dynamics has been shown to play an important role in the diagnosis of ischemic coronary artery disease. Accurate extraction of complex coronary artery structures in a watertight geometry is a prerequisite, but manual segmentation is both tedious and subjective. Several semi- and fully automated coronary artery extraction approaches have been developed but have faced several challenges. Conventional voxel-based methods allow for watertight segmentation but are slow and difficult to incorporate expert knowledge. Machine learning based methods are relatively fast and capture rich information embedded in manual annotations. Although sufficient for visualization and analysis of coronary anatomy, these methods cannot be used directly for blood flow simulation if the coronary vasculature is represented as a loose combination of tubular structures and the bifurcation geometry is improperly modeled. In this paper, we propose a novel method to extract branching coronary arteries from CT imaging with a focus on explicit bifurcation modeling and application of machine learning. A bifurcation lumen is firstly modeled by generating the convex hull to join tubular vessel branches. Guided by the pre-determined centerline, machine learning based segmentation is performed to adapt the bifurcation lumen model to target vessel boundaries and smoothed by subdivision surfaces. Our experiments show the constructed coronary artery geometry from CT imaging is accurate by comparing results against the manually annotated ground-truths, and can be directly applied to coronary blood flow simulation.

Effect of Varying Hemodynamic and Vascular Conditions on Fractional Flow Reserve: An In Vitro Study.

BACKGROUND: The aim of this study was to investigate the impact of varying hemodynamic conditions on fractional flow reserve (ratio of pressure distal [Pd] and proximal [Pa] to stenosis under hyperemia) in an in vitro setting. Failure to achieve maximal hyperemia and the choice of hyperemic agents may have differential effects on coronary hemodynamics and, consequently, on the determination of fractional flow reserve. METHODS AND RESULTS: An in vitro flow system was developed to experimentally model the physiological coronary circulation as flow-dependent stenosis resistance in series with variable downstream resistance. Five idealized models with 30% to 70% diameter stenosis severity were fabricated using VeroClear rigid material in an Objet260 Connex printer. Mean aortic pressure was maintained at 7 levels (60-140 mm Hg) from hypotension to hypertension using a needle valve that mimicked adjustable microcirculatory resistance. A range of physiological flow rates was applied by a steady flow pump and titrated by a flow sensor. The pressure drop and the pressure ratio (Pd/Pa) were assessed for the 7 levels of aortic pressure and differing flow rates. The in vitro experimental data were coupled with pressure-flow relationships from clinical data for populations with and without myocardial infarction, respectively, to evaluate fractional flow reserve. The curve for pressure ratio and flow rate demonstrated a quadratic relationship with a decreasing slope. The absolute decrease in fractional flow reserve in the group without myocardial infarction (with myocardial infarction) was on the order of 0.03 (0.02), 0.05 (0.02), 0.07 (0.05), 0.17 (0.13) and 0.20 (0.24), respectively, for 30%, 40%, 50%, 60%, and 70% diameter stenosis, for an increase in aortic pressure from 60 to 140 mm Hg. CONCLUSIONS: The fractional flow reserve value, an index of physiological stenosis significance, was observed to decrease with increasing aortic pressure for a given stenosis in this idealized in vitro experiment for vascular groups with and without myocardial infarction.

Using edge-preserving algorithm with non-local mean for significantly improved image-domain material decomposition in dual-energy CT.

Increased noise is a general concern for dual-energy material decomposition. Here, we develop an image-domain material decomposition algorithm for dual-energy CT (DECT) by incorporating an edge-preserving filter into the Local HighlY constrained backPRojection reconstruction (HYPR-LR) framework. With effective use of the non-local mean, the proposed algorithm, which is referred to as HYPR-NLM, reduces the noise in dual-energy decomposition while preserving the accuracy of quantitative measurement and spatial resolution of the material-specific dual-energy images. We demonstrate the noise reduction and resolution preservation of the algorithm with an iodine concentrate numerical phantom by comparing the HYPR-NLM algorithm to the direct matrix inversion, HYPR-LR and iterative image-domain material decomposition (Iter-DECT). We also show the superior performance of the HYPR-NLM over the existing methods by using two sets of cardiac perfusing imaging data. The DECT material decomposition comparison study shows that all four algorithms yield acceptable quantitative measurements of iodine concentrate. Direct matrix inversion yields the highest noise level, followed by HYPR-LR and Iter-DECT. HYPR-NLM in an iterative formulation significantly reduces image noise and the image noise is comparable to or even lower than that generated using Iter-DECT. For the HYPR-NLM method, there are marginal edge effects in the difference image, suggesting the high-frequency details are well preserved. In addition, when the search window size increases from to , there are no significant changes or marginal edge effects in the HYPR-NLM difference images. The reference drawn from the comparison study includes: (1) HYPR-NLM significantly reduces the DECT material decomposition noise while preserving quantitative measurements and high-frequency edge information, and (2) HYPR-NLM is robust with respect to parameter selection.

Influence of virtual intervention and blood rheology on mass transfer through thoracic aortic aneurysm.

Computational fluid dynamics tools have been used to investigate blood flow through the human thoracic aortic models with aneurysm before and after virtual stent graft operation. The impact of blood rheology and aortic geometry on the wall shear stress (WSS), luminal surface low-density lipoproteins (LDL) concentration, and oxygen flux along the arterial wall is investigated. The stent graft at the aneurysm has significant effects on WSS and mass transport in blood flow. Due to the low flow rate, Newtonian blood assumption generally under-estimates the WSS. The non-Newtonian blood rheology play an important role in the LDL transport as well as oxygen transport. It is found that WSS alone is insufficient to correctly predict the location with high risk of atherogenesis. The results suggest that WSS, luminal surface LDL concentration, and the oxygen flux on the wall have to be considered together to evaluate the performance of virtual operation.

Myocardial perfusion analysis in cardiac computed tomography angiographic images at rest.

Cardiac computed tomography angiography (CTA) is a non-invasive method for anatomic evaluation of coronary artery stenoses. However, CTA is prone to artifacts that reduce the diagnostic accuracy to identify stenoses. Further, CTA does not allow for determination of the physiologic significance of the visualized stenoses. In this paper, we propose a new system to determine the physiologic manifestation of coronary stenoses by assessment of myocardial perfusion from typically acquired CTA images at rest. As a first step, we develop an automated segmentation method to delineate the left ventricle. Both endocardium and epicardium are compactly modeled with subdivision surfaces and coupled by explicit thickness representation. After initialization with five anatomical landmarks, the model is adapted to a target image by deformation increments including control vertex displacements and thickness variations guided by trained AdaBoost classifiers, and regularized by a prior of deformation increments from principal component analysis (PCA). The evaluation using a 5-fold cross-validation demonstrates the overall segmentation error to be 1.00 ± 0.39 mm for endocardium and 1.06 ± 0.43 mm for epicardium, with a boundary contour alignment error of 2.79 ± 0.52. Based on our LV model, two types of myocardial perfusion analyzes have been performed. One is a perfusion network analysis, which explores the correlation (as network edges) pattern of perfusion between all pairs of myocardial segments (as network nodes) defined in AHA 17-segment model. We find perfusion network display different patterns in the normal and disease groups, as divided by whether significant coronary stenosis is present in quantitative coronary angiography (QCA). The other analysis is a clinical validation assessment of the ability of the developed algorithm to predict whether a patient has significant coronary stenosis when referenced to an invasive QCA ground truth standard. By training three machine learning techniques using three features of normalized perfusion intensity, transmural perfusion ratio, and myocardial wall thickness, we demonstrate AdaBoost to be slightly better than Naive Bayes and Random Forest by the area under receiver operating characteristics (ROC) curve. For the AdaBoost algorithm, an optimal cut-point reveals an accuracy of 0.70, with sensitivity and specificity of 0.79 and 0.64, respectively. Our study shows perfusion analysis from CTA images acquired at rest is useful for providing physiologic information in diagnosis of obstructive coronary artery stenoses.

Myocardial scar imaging by standard single-energy and dual-energy late enhancement CT: Comparison with pathology and electroanatomic map in an experimental chronic infarct porcine model.

BACKGROUND: Myocardial scar is a substrate for ventricular tachycardia and sudden cardiac death. Late enhancement CT imaging can detect scar, but it remains unclear whether newer late enhancement dual-energy (LE-DECT) acquisition has benefit over standard single-energy late enhancement (LE-CT). OBJECTIVE: We aim to compare late enhancement CT using newer LE-DECT acquisition and single-energy LE-CT acquisitions with pathology and electroanatomic map (EAM) in an experimental chronic myocardial infarction (MI) porcine study. METHODS: In 8 pigs with chronic myocardial infarction (59 ± 5 kg), we performed dual-source CT, EAM, and pathology. For CT imaging, we performed 3 acquisitions at 10 minutes after contrast administration: LE-CT 80 kV, LE-CT 100 kV, and LE-DECT with 2 postprocessing software settings. RESULTS: Of the sequences, LE-CT 100 kV provided the best contrast-to-noise ratio (all P ≤ .03) and correlation to pathology for scar (ρ = 0.88). LE-DECT overestimated scar (both P = .02), whereas LE-CT images did not (both P = .08). On a segment basis (n = 136), all CT sequences had high specificity (87%-93%) and modest sensitivity (50%-67%), with LE-CT 100 kV having the highest specificity of 93% for scar detection compared to pathology and agreement with EAM (κ = 0.69). CONCLUSIONS: Standard single-energy LE-CT, particularly 100 kV, matched better to pathology and EAM than dual-energy LE-DECT for scar detection. Larger human trials as well as more technical studies that optimize varying different energies with newer hardware and software are warranted.

Current progress in 3D printing for cardiovascular tissue engineering.

3D printing is a technology that allows the fabrication of structures with arbitrary geometries and heterogeneous material properties. The application of this technology to biological structures that match the complexity of native tissue is of great interest to researchers. This mini-review highlights the current progress of 3D printing for fabricating artificial tissues of the cardiovascular system, specifically the myocardium, heart valves, and coronary arteries. In addition, how 3D printed sensors and actuators can play a role in tissue engineering is discussed. To date, all the work with building 3D cardiac tissues have been proof-of-principle demonstrations, and in most cases, yielded products less effective than other traditional tissue engineering strategies. However, this technology is in its infancy and therefore there is much promise that through collaboration between biologists, engineers and material scientists, 3D bioprinting can make a significant impact on the field of cardiovascular tissue engineering.

Heart-rate dependent improvement in image quality and diagnostic accuracy of coronary computed tomographic angiography by novel intracycle motion correction algorithm.

BACKGROUND: To determine the effect of a novel intracycle motion correction algorithm (MCA) on diagnostic accuracy of coronary computed tomographic angiography. METHODS: Coronary artery phantom models were scanned at static and heart rate (HR) simulation of 60-100 beat/min and reconstructed with a conventional algorithm and MCA. RESULTS: Among 144 coronary segments, improvements in image interpretability, quality, and diagnostic accuracy by MCA were observed for HRs of 80 and 100 (P<.05 for all), but not for HR of 60. CONCLUSION: Novel intracycle MCA demonstrates improved HR-dependent image interpretability, and quality and accuracy, particularly at higher HRs.

Analysis of ventricular function by CT.

The assessment of ventricular function, cardiac chamber dimensions, and ventricular mass is fundamental for clinical diagnosis, risk assessment, therapeutic decisions, and prognosis in patients with cardiac disease. Although cardiac CT is a noninvasive imaging technique often used for the assessment of coronary artery disease, it can also be used to obtain important data about left and right ventricular function and morphology. In this review, we will discuss the clinical indications for the use of cardiac CT for ventricular analysis, review the evidence on the assessment of ventricular function compared with existing imaging modalities such cardiac magnetic resonance imaging and echocardiography, provide a typical cardiac CT protocol for image acquisition and postprocessing for ventricular analysis, and provide step-by-step instructions to acquire multiplanar cardiac views for ventricular assessment from the standard axial, coronal, and sagittal planes. Furthermore, both qualitative and quantitative assessments of ventricular function as well as sample reporting are detailed.

Registration of coronary arteries in computed tomography angiography images using Hidden Markov Model.

Computed tomography angiography (CTA) allows for not only diagnosis of coronary artery disease (CAD) with high spatial resolution but also monitoring the remodeling of vessel walls in the progression of CAD. Alignment of coronary arteries in CTA images acquired at different times (with a 3-7 years interval) is required to visualize and analyze the geometric and structural changes quantitatively. Previous work in image registration primarily focused on large anatomical structures and leads to suboptimal results when applying to registration of coronary arteries. In this paper, we develop a novel method to directly align the straightened coronary arteries in the cylindrical coordinate system guided by the extracted centerlines. By using a Hidden Markov Model (HMM), image intensity information from CTA and geometric information of extracted coronary arteries are combined to align coronary arteries. After registration, the pathological features in two straightened coronary arteries can be directly visualized side by side by synchronizing the corresponding cross-sectional slices and circumferential rotation angles. By evaluating with manually labeled landmarks, the average distance error is 1.6 mm.

Methods for characterizing human coronary artery deformation from cardiac-gated computed tomography data.

Accurate quantification of changes in length, curvature, and bifurcation angles of coronary arteries due to cardiac motion is important for the design of coronary stents. A new method is developed to describe the dynamic characteristics of the human coronary artery. From cardiac-gated computed tomography (CT) data, 3-D surface geometry and centerline paths of the coronary arteries were constructed. For quantification of strain and twisting deformation, 3-D distortion-free vessel straightening and landmark matching algorithms were developed to compute the relative translation and rotation of distal landmarks with respect to a proximal landmark. For quantification of bending deformation, change in curvature was measured by computing a best-fit torus in the region of interest within a coronary segment. The optimal torus parameters were estimated by minimizing the standard deviation of distances from the surface mesh to the centerline of the torus. The angle between branch vessels was measured using linear fitting of centroid sets from the cross-sectional vessel lumen. The proposed methods were verified using a software phantom and applied to two patient specific CT datasets. Vascular deformations derived from these methods can provide information for designing bench-top tests for endovascular devices that better replicate the in vivo environment, thereby improving device performance prediction and leading to more durable designs.

Tracking the motion trajectories of junction structures in 4D CT images of the lung.

Respiratory motion poses a major challenge in lung radiotherapy. Based on 4D CT images, a variety of intensity-based deformable registration techniques have been proposed to study the pulmonary motion. However, the accuracy achievable with these approaches can be sub-optimal because the deformation is defined globally in space. Therefore, the accuracy of the alignment of local structures may be compromised. In this work, we propose a novel method to detect a large collection of natural junction structures in the lung and use them as the reliable markers to track the lung motion. Specifically, detection of the junction centers and sizes is achieved by analysis of local shape profiles on one segmented image. To track the temporal trajectory of a junction, the image intensities within a small region of interest surrounding the center are selected as its signature. Under the assumption of the cyclic motion, we describe the trajectory by a closed B-spline curve and search for the control points by maximizing a metric of combined correlation coefficients. Local extrema are suppressed by improving the initial conditions using random walks from pair-wise optimizations. Several descriptors are introduced to analyze the motion trajectories. Our method was applied to 13 real 4D CT images. More than 700 junctions in each case are detected with an average positive predictive value of greater than 90%. The average tracking error between automated and manual tracking is sub-voxel and smaller than the published results using the same set of data.

Virtual Interventions for Image-based Blood Flow Computation.

Image-based blood flow computation provides great promise for evaluation of vascular devices and assessment of surgical procedures. However, many previous studies employ idealized arterial and device models or only patient-specific models from the image data after device deployment, since the tools for model construction are unavailable or limited and tedious to use. Moreover, in contrast to retrospective studies from existing data, there is a pressing need for prospective analysis with the goal of surgical planning. Therefore, it is necessary to construct models with deployed devices in a fast, virtual and interactive fashion. The goal of this paper is to develop new geometric methods to deploy stents or stent grafts virtually to patient-specific geometric models constructed from a 3D segmentation of medical images. A triangular surface representing the vessel lumen boundary is extracted from the segmentation. The diseased portion is either clipped and replaced by the surface of a deployed device or rerouted in the case of a bypass graft. For diseased arteries close to bifurcations, bifurcated device models are generated. A method to map a 2D strut pattern on the surface of a device is also presented. We demonstrate three applications of our methods in personalized surgical planning for aortic aneurysms, aortic coarctation, and coronary artery stenosis using blood flow computation. Our approach enables prospective model construction and may help to expand the throughput required by routine clinical uses in the future.

Deformable image registration of liver with consideration of lung sliding motion.

PURPOSE: A feature based deformable registration model with sliding transformation was developed in the upper abdominal region for liver cancer. METHODS: A two-step thin-plate spline (bi-TPS) algorithm was implemented to deformably register the liver organ. The first TPS registration was performed to exclusively quantify the sliding displacement component. A manual segmentation of the thoracic and abdominal cavity was performed as a priori knowledge. Tissue feature points were automatically identified inside the segmented contour on the images. The scale invariant feature transform method was utilized to match feature points that served as landmarks for the subsequent TPS registration to derive the sliding displacement vector field. To a good approximation, only motion along superior/inferior (SI) direction of voxels on each slice was averaged to obtain the sliding displacement for each slice. A second TPS transformation, as the last step, was carried out to obtain the local deformation field. Manual identification of bifurcation on liver, together with the manual segmentation of liver organ, was employed as a "ground truth" for assessing the algorithm's performance. RESULTS: The proposed two-step TPS was assessed with six liver patients. The average error of liver bifurcation between manual identification and calculation for these patients was less than 1.8 mm. The residual errors between manual contour and propagated contour of liver organ using the algorithm fell in the range between 2.1 and 2.8 mm. An index of Dice similarity coefficient (DSC) between manual contour and calculated contour for liver tumor was 93.6% compared with 71.2% from the conventional TPS calculation. CONCLUSIONS: A high accuracy (∼2 mm) of the two-step feature based TPS registration algorithm was achievable for registering the liver organ. The discontinuous motion in the upper abdominal region was properly taken into consideration. Clinical implementation of the algorithm will find broad application in radiation therapy of liver cancer.

Automated detection of junctions structures and tracking of their trajectories in 4D images.

Junction structures, as the natural anatomical markers, are useful to study the organ or tumor motion. However, detection and tracking of the junctions in four-dimensional (4D) images are challenging. The paper presents a novel framework to automate this task. Detection of their centers and sizes is first achieved by an analysis of local shape profiles on one segmented reference image. Junctions are then separately tracked by simultaneously using neighboring intensity features from all images. Defined by a closed B-spline space curve, the individual trajectory is assumed to be cyclic and obtained by maximizing the metric of combined correlation coefficients. Local extrema are suppressed by improving the initial conditions using random walks from pair-wise optimizations. Our approach has been applied to analyze the vessel junctions in five real 4D respiration-gated computed tomography (CT) image datasets with promising results. More than 500 junctions in the lung are detected with an average accuracy of greater than 85% and the mean error between the automated and the manual tracking is sub-voxel.

Simulation of blood flow in deformable vessels using subject-specific geometry and spatially varying wall properties.

Simulation of blood flow using image-based models and computational fluid dynamics has found widespread application to quantifying hemodynamic factors relevant to the initiation and progression of cardiovascular diseases and for planning interventions. Methods for creating subject-specific geometric models from medical imaging data have improved substantially in the last decade but for many problems, still require significant user interaction. In addition, while fluid-structure interaction methods are being employed to model blood flow and vessel wall dynamics, tissue properties are often assumed to be uniform. In this paper, we propose a novel workflow for simulating blood flow using subject-specific geometry and spatially varying wall properties. The geometric model construction is based on 3D segmentation and geometric processing. Variable wall properties are assigned to the model based on combining centerline-based and surface-based methods. We finally demonstrate these new methods using an idealized cylindrical model and two subject-specific vascular models with thoracic and cerebral aneurysms.

Physics-based modeling of aortic wall motion from ECG-gated 4D computed tomography.

Recent advances in electrocardiogram (ECG)-gated Computed Tomography (CT) technology provide 4D (3D+T) information of aortic wall motion in high spatial and temporal resolution. However, imaging artifacts, e.g. noise, partial volume effect, misregistration and/or motion blurring may preclude its usability in many applications where accuracy and reliability are concerns. Although it is possible to find correspondence through tagged MRI or echo or image registration, it may be either inconsistent to the physics or difficult to utilize data from all frames. In this paper, we propose a physics-based filtering approach to construct a dynamic model from these 4D images. It includes a state filter that corrects simulated displacements from an elastic finite element model to match observed motion from images. In the meantime, the model parameters are refined to improve the model quality by applying a parameter filter based on ensemble Kalman filtering. We evaluated the performance of our method on synthetic data where ground-truths are available. Finally, we successfully applied the method to a real data set.

Virtual stent grafting in personalized surgical planning for treatment of aortic aneurysms using image-based computational fluid dynamics.

Image-based computational fluid dynamics provides great promise for evaluation of vascular devices and assessment of surgical procedures. However, many previous studies employ idealized arterial and device models or patient-specific models with a limited number of cases, since the model construction process is tedious and time-consuming. Moreover, in contrast to retrospective studies from existing image data, there is a pressing need of prospective analysis with the goal of surgical planning. Therefore, it is necessary to construct models with implanted devices in a fast, virtual and interactive fashion. The goal of this paper is to develop new geometric methods to deploy stent grafts virtually to patient-specific models constructed from direct 3D segmentation of medical images. A triangular surface representing vessel lumen boundary is extracted from the segmentation. The diseased portion is then clipped and replaced by the surface of a virtual stent graft following the centerline obtained from the clipped portion. A Y-shape stent graft is employed in case of bifurcated arteries. A method to map a 2D strut pattern on the stent graft is also presented. We demonstrate the application of our methods to quantify wall shear stresses and forces acting on stent grafts in personalized surgical planning for endovascular treatment of thoracic and abdominal aortic aneurysms. Our approach enables prospective model construction and may help to increase its throughput required by routine clinical uses in the future.

Automated neurite labeling and analysis in fluorescence microscopy images.

BACKGROUND: To investigate the intricate nervous processes involved in many biological activities by computerized image analysis, accurate and reproducible labeling and measurement of neurites are prerequisite. We have developed an automated neurite analysis method to assist this task. METHODS: Our approach can be considered as automated with certain user interaction in setting initial parameters. Single and connected centerlines along neurites are extracted. The computerized method can also generate branching and end points. Owing to its multi-scale flexibility, both thick and thin neurites are simultaneously detected. RESULTS: We employ the relative neurite length difference (defined as the difference between the lengths obtained by automated and manual analysis divided by the total length of the latter) and neurite centerline deviation (defined as the area of the regions enclosed by different paths between automated and manual analysis divided by the total length of the former) to evaluate the performance of our algorithm, which is of great interest in neurite analysis. The average of the relative length difference is about 0.02, while the average of the centerline deviation is about 2.8 pixels. The probabilities of the distributions being the same from the Kolmogorov-Smirnov (KS) test of the automatic and manual results are 99.79%. The KS test also shows no significant bias between different observers based on the proposed new validation scheme. CONCLUSIONS: With the accurate and automated extraction of neurite centerlines and measurement of neurite lengths, the proposed method, which greatly reduces human labor and improves efficiency, can serve as a candidate tool for large-scale neurite analysis beyond the capability of manual tracing methods.