Over the years, three-dimensional models of the mitral valve have generally been organized around a simplified anatomy. Leaflets have been typically modeled as membranes, tethered to discrete chordae typically modeled as one-dimensional, non-linear cables. Yet, recent, high-resolution medical images have revealed that there is no clear boundary between the chordae and the leaflets. In fact, the mitral valve has been revealed to be more of a webbed structure whose architecture is continuous with the chordae and their extensions into the leaflets. Such detailed images can serve as the basis of anatomically accurate, subject-specific models, wherein the entire valve is modeled with solid elements that more faithfully represent the chordae, the leaflets, and the transition between the two. These models have the potential to enhance our understanding of mitral valve mechanics and to re-examine the role of the mitral valve chordae, which heretofore have been considered to be ‘invisible’ to the fluid and to be of secondary importance to the leaflets. However, these new models also require a rethinking of modeling assumptions. In this study, we examine the conventional practice of loading the leaflets only and not the chordae in order to study the structural response of the mitral valve apparatus. Specifically, we demonstrate that fully resolved 3D models of the mitral valve require a fluid–structure interaction analysis to correctly load the valve even in the case of quasi-static mechanics. While a fluid–structure interaction mode is still more computationally expensive than a structural-only model, we also show that advances in GPU computing have made such models tractable.
Numerical models of the mitral valve have been used to elucidate mitral valve function and mechanics. These models have evolved from simple two-dimensional approximations to complex three-dimensional fully coupled fluid structure interaction models. However, to date these models lack direct one-to-one experimental validation. As computational solvers vary considerably, experimental benchmark data are critically important to ensure model accuracy. In this study, a novel left heart simulator was designed specifically for the validation of numerical mitral valve models. Several distinct experimental techniques were collectively performed to resolve mitral valve geometry and hemodynamics. In particular, micro-computed tomography was used to obtain accurate and high-resolution (39 μm voxel) native valvular anatomy, which included the mitral leaflets, chordae tendinae, and papillary muscles. Three-dimensional echocardiography was used to obtain systolic leaflet geometry. Stereoscopic digital particle image velocimetry provided all three components of fluid velocity through the mitral valve, resolved every 25 ms in the cardiac cycle. A strong central filling jet (V ∼ 0.6 m/s) was observed during peak systole with minimal out-of-plane velocities. In addition, physiologic hemodynamic boundary conditions were defined and all data were synchronously acquired through a central trigger. Finally, the simulator is a precisely controlled environment, in which flow conditions and geometry can be systematically prescribed and resultant valvular function and hemodynamics assessed. Thus, this work represents the first comprehensive database of high fidelity experimental data, critical for extensive validation of mitral valve fluid structure interaction simulations.
We propose a simple method for reconstructing vascular trees from 3D images. Our algorithm extracts persistent maxima of the intensity on all axis-aligned 2D slices of the input image. The maxima concentrate along 1D intensity ridges, in particular along blood vessels. We build a forest connecting the persistent maxima with short edges. The forest tends to approximate the blood vessels present in the image, but also contains numerous spurious features and often fails to connect segments belonging to one vessel in low contrast areas. We improve the forest by applying simple geometric filters that trim short branches, fill gaps in blood vessels and remove spurious branches from the vascular tree to be extracted. Experiments show that our technique can be applied to extract coronary trees from heart CT scans.
Computed tomography (CT) slices are combined with computational fluid dynamics (CFD) to simulate the flow patterns in a human left coronary artery. The vascular model was reconstructed from CT slices scanned from a healthy volunteer in vivo. The spatial resolution of the slices is 0.6 × 0.6 × 0.625 mm so that geometrical details of the local wall surface of the vessel could be considered in the CFD modeling. This level of resolution is needed to investigate the wall shear stress (WSS) distribution, a factor generally recognized as a related to the atherogenesis. The WSS distributions on the main trunk and bifurcation of the left coronary artery of the model in one cardiac cycle are presented, and the results demonstrate that low and oscillating WSS is correlative with clinical observations of the atherosclerotic-prone sites in the left coronary artery.
Patient-specific models of the heart's mitral valve (MV) exhibit potential for surgical planning. While advances in 3D echocardiography (3DE) have provided adequate resolution to extract MV leaflet geometry, no study has quantitatively assessed the accuracy of their modeled leaflets vs. a ground-truth standard for temporal frames beyond systolic closure or for differing valvular dysfunctions. The accuracy of a 3DE-based segmentation methodology based on J-splines was assessed for porcine MVs with known 4D leaflet coordinates within a pulsatile simulator during closure, peak closure, and opening for a control, prolapsed, and billowing MV model. For all time points, the mean distance error between the segmented models and ground-truth data were 0.40 ± 0.32 mm, 0.52 ± 0.51 mm, and 0.74 ± 0.69 mm for the control, flail, and billowing models. For all models and temporal frames, 95% of the distance errors were below 1.64 mm. When applied to a patient data set, segmentation was able to confirm a regurgitant orifice and post-operative improvements in coaptation. This study provides an experimental platform for assessing the accuracy of an MV segmentation methodology at phases beyond systolic closure and for differing MV dysfunctions. Results demonstrate the accuracy of a MV segmentation methodology for the development of future surgical planning tools.
Motivation: Virus phylogeographers rely on DNA sequences of viruses and the locations of the infected hosts found in public sequence databases like GenBank for modeling virus spread. However, the locations in GenBank records are often only at the country or state level, and may require phylogeographers to scan the journal articles associated with the records to identify more localized geographic areas. To automate this process, we present a named entity recognizer (NER) for detecting locations in biomedical literature. We built the NER using a deep feedforward neural network to determine whether a given token is a toponym or not. To overcome the limited human annotated data available for training, we use distant supervision techniques to generate additional samples to train our NER. Results: Our NER achieves an F1-score of 0.910 and significantly outperforms the previous stateof- the-art system. Using the additional data generated through distant supervision further boosts the performance of the NER achieving an F1-score of 0.927. The NER presented in this research improves over previous systems significantly. Our experiments also demonstrate the NER?s capability to embed external features to further boost the system?s performance. We believe that the same methodology can be applied for recognizing similar biomedical entities in scientific literature.
BACKGROUND AND PURPOSE: Phase analysis of gated single-photon emission computed tomography (SPECT) myocardial perfusion imaging (MPI) has been validated as a reliable tool to assess left-ventricular (LV) mechanical dyssynchrony. The initial results were all confirmed from studies using technetium-99m (Tc-99m) sestamibi or tetrofosmin as the radiotracers. The purpose of this study was to evaluate the feasibility of phase analysis in thallium-201 (Tl-201) gated SPECT MPI. MATERIALS AND METHODS: Seventeen patients referred from a cardiology clinic for evaluation of coronary artery disease were studied. All patients underwent both Tl-201 and Tc-99m sestamibi gated SPECT MPI within 1 week. An additional 34 patients with Tl-201 gated SPECT and 22 patients with Tc-99m sestamibi gated SPECT, who had a low likelihood of coronary artery disease, normal LV function, and normal perfusion on MPI, were used as normal controls. LV dyssynchrony parameters, including phase standard deviation (PSD) and phase histogram bandwidth (PHB), were measured using a standard phase analysis tool and compared between Tl-201 and Tc-99m sestamibi images. RESULTS: The LV dyssynchrony parameters correlated well (r=0.93 for PSD and r=0.84 for PHB) between Tl-201 and Tc-99m sestamibi images. The dyssynchrony parameters of Tl-201 were significantly larger than those of Tc-99m sestamibi (PSD: 24.5±12.0 vs. 17.4±9.7, P<0.001; PHB: 74.7±35.5 vs. 50.6±25.0, P<0.001). In comparison with normal controls, Tl-201 and Tc-99m sestamibi images showed concordant results. CONCLUSION: LV dyssynchrony parameters correlated well between Tl-201 and Tc-99m sestamibi images, even though the values were significantly larger for Tl-201 than for Tc-99m sestamibi. Tl-201 images showed results similar to those of Tc-99m sestamibi in the diagnosis of LV dyssynchrony.
Total cavopulmonary connection is the result of a series of palliative surgical repairs performed on patients with single ventricle heart defects. The resulting anatomy has complex and unsteady hemodynamics characterized by flow mixing and flow separation. Although varying degrees of flow pulsatility have been observed in vivo, non-pulsatile (time-averaged) boundary conditions have traditionally been assumed in hemodynamic modeling, and only recently have pulsatile conditions been incorporated without completely characterizing their effect or importance. In this study, 3D numerical simulations with both pulsatile and non-pulsatile boundary conditions were performed for 24 patients with different anatomies and flow boundary conditions from Georgia Tech database. Flow structures, energy dissipation rates and pressure drops were compared under rest and simulated exercise conditions. It was found that flow pulsatility is the primary factor in determining the appropriate choice of boundary conditions, whereas the anatomic configuration and cardiac output had secondary effects. Results show that the hemodynamics can be strongly influenced by the presence of pulsatile flow. However, there was a minimum pulsatility threshold, identified by defining a weighted pulsatility index (wPI), above which the influence was significant. It was shown that when wPI < 30%, the relative error in hemodynamic predictions using time-averaged boundary conditions was less than 10% compared to pulsatile simulations. In addition, when wPI < 50, the relative error was less than 20%. A correlation was introduced to relate wPI to the relative error in predicting the flow metrics with non-pulsatile flow conditions.
Pluripotent embryonic stem cells (ESCs) have the unique ability to differentiate into cells from all germ lineages, making them a potentially robust cell source for regenerative medicine therapies, but difficulties in predicting and controlling ESC differentiation currently limit the development of therapies and applications from such cells. A common approach to induce the differentiation of ESCs in vitro is via the formation of multicellular aggregates known as embryoid bodies (EBs), yet cell fate specification within EBs is generally considered an ill-defined and poorly controlled process. Thus, the objective of this study was to use rules-based cellular modeling to provide insight into which processes influence initial cell fate transitions in 3-dimensional microenvironments. Mouse embryonic stem cells (D3 cell line) were differentiated to examine the temporal and spatial patterns associated with loss of pluripotency as measured through Oct4 expression. Global properties of the multicellular aggregates were accurately recapitulated by a physics-based aggregation simulation when compared to experimentally measured physical parameters of EBs. Oct4 expression patterns were analyzed by confocal microscopy over time and compared to simulated trajectories of EB patterns. The simulations demonstrated that loss of Oct4 can be modeled as a binary process, and that associated patterns can be explained by a set of simple rules that combine baseline stochasticity with intercellular communication. Competing influences between Oct4+ and Oct4- neighbors result in the observed patterns of pluripotency loss within EBs, establishing the utility of rules-based modeling for hypothesis generation of underlying ESC differentiation processes. Importantly, the results indicate that the rules dominate the emergence of patterns independent of EB structure, size, or cell division. In combination with strategies to engineer cellular microenvironments, this type of modeling approach is a powerful tool to predict stem cell behavior under a number of culture conditions that emulate characteristics of 3D stem cell niches.
Computational models of the heart's mitral valve (MV) exhibit potential for preoperative surgical planning in ischemic mitral regurgitation (IMR). However challenges exist in defining boundary conditions to accurately model the function and response of the chordae tendineae to both IMR and surgical annuloplasty repair. Towards this goal, a ground-truth data set was generated by quantifying the isolated effects of IMR and mitral annuloplasty on leaflet coaptation, regurgitation, and tethering forces of the anterior strut and posterior intermediary chordae tendineae. MVs were excised from ovine hearts (N = 15) and mounted in a pulsatile heart simulator which has been demonstrated to mimic the systolic MV geometry and coaptation of healthy and chronic IMR sheep. Strut and intermediary chordae from both MV leaflets (N = 4) were instrumented with force transducers. Tested conditions included a healthy control, IMR, oversized annuloplasty, true-sized annuloplasty, and undersized mitral annuloplasty. A2-P2 leaflet coaptation length, regurgitation, and chordal tethering were quantified and statistically compared across experimental conditions. MR was successfully simulated with significant increases in MR, tethering forces for each of the chordae, and decrease in leaflet coaptation (p < .05). Compared to the IMR condition, increasing levels of downsized annuloplasty significantly reduced regurgitation, increased coaptation, reduced posteromedial papillary muscle strut chordal forces, and reduced intermediary chordal forces from the anterolateral papillary muscle (p < .05). These results provide for the first time a novel comprehensive data set for refining the ability of computational MV models to simulate IMR and varying sizes of complete rigid ring annuloplasty.