Ben A. Lin

Color Doppler jet area overestimates regurgitant volume when multiple jets are present

BACKGROUND Color Doppler jet area (CDJA) is an important measure used to classify mitral regurgitation (MR) severity. The investigators hypothesized that the presence and configuration of multiple regurgitant jets can alter CDJA quantification for fixed regurgitant volumes. This has relevance to MR assessment prior to the treatment of valves with multiple regurgitant orifices or after surgical or percutaneous double-orifice mitral valve repair. METHODS An in vitro model was developed to create jets flowing through a simulated mitral orifice into an imaging chamber. The flow loop was driven with a pulsatile pump at 60 beats/min containing a water-glycerol solution approximating the viscosity of blood. At the orifice, simulated regurgitant stroke volumes of 2.5 to 25 mL were created through plates having either single openings with orifice areas from 0.125 to 0.50 cm(2) or two to four openings with total orifice area of 0.25 cm(2) and varied linear spacing. An 8-MHz transthoracic two-dimensional ultrasound probe was used to acquire jet velocities by continuous-wave Doppler as well as color Doppler for offline analysis. CDJA values were obtained with custom automated pixel-counting software. RESULTS Peak jet velocities ranged from 30 to 550 cm/sec. For single jets, normalized average CDJA values increased nonlinearly as a function of average Reynolds number. Peak CDJA values were up to 62% higher for multiple jets compared with single jets with similar total orifice areas and simulated regurgitant stroke volumes. The presence or absence of multiple jets, rather than the total number of jets, appeared to have a greater effect on maximum CDJA. In addition, peak CDJA values for multiple jets increased with increased linear spacing. CONCLUSIONS A fixed regurgitant volume involving multiple jets will have a larger CDJA value than the same total volume from a single jet. The source of this discrepancy appears to be increased ambient fluid entrainment from adjacent regurgitant jets. This potential overestimation of MR severity using color Doppler flow jets should be taken into consideration when assessing MR prior to treatment or when assessing residual MR after double-orifice mitral valve repair.

Clinical Application of Three-Dimensional Echocardiography

Echocardiography is one of the most valuable diagnostic tools in cardiology. Technological advances in ultrasound, computer and electronics enables three-dimensional (3-D) imaging to be a clinically viable modality which has significant impact on diagnosis, management and interventional procedures. Since the inception of 3D fully-sampled matrix transthoracic and transesophageal technology it has enabled easier acquisition, immediate on-line display, and availability of on-line analysis for the left ventricle, right ventricle and mitral valve. The use of 3D TTE has mainly focused on mitral valve disease, left and right ventricular volume and functional analysis. As structural heart disease procedures become more prevalent, 3D TEE has become a requirement for preparation of the procedure, intra-procedural guidance as well as monitoring for complications and device function. We anticipate that there will be further software development, improvement in image quality and workflow.

Segmentation of left ventricles from echocardiographic sequences via sparse appearance representation

Sparse representation has proven to be a powerful mathematical framework for studying high-dimensional data and uncovering its structures. Some recent research has shown its promise in discriminating image patterns. This paper presents an approach employing sparse appearance representation for segmenting left ventricular endocardial and epicardial boundaries from 2D echocardiographic sequences. It leverages the inherent spatio-temporal coherence of tissue/blood appearance over the sequence by modeling the different appearance of blood and tissues with different appearance dictionaries and updating the dictionaries in a boosting framework as the frames are segmented sequentially. The appearance of each frame is predicted in the form of appearance dictionaries based on the appearance observed in the preceding frames. The dictionaries discriminate image patterns by reconstructing them in the process of sparse coding resulting in an appearance discriminant that we incorporate into a region-based level set segmentation process. We illustrate the advantages of our approach by comparing it to manual tracings and an intensity-prior-based level set method. Experimental results on 34 2D canine echocardiographic sequences show that sparse appearance representation significantly outperforms intensity in terms of reliability and accuracy of segmentation.

Towards patient-specific modeling of mitral valve repair: 3D transesophageal echocardiography-derived parameter estimation

&NA; Transesophageal echocardiography (TEE) is routinely used to provide important qualitative and quantitative information regarding mitral regurgitation. Contemporary planning of surgical mitral valve repair, however, still relies heavily upon subjective predictions based on experience and intuition. While patient‐specific mitral valve modeling holds promise, its effectiveness is limited by assumptions that must be made about constitutive material properties. In this paper, we propose and develop a semi‐automated framework that combines machine learning image analysis with geometrical and biomechanical models to build a patient‐specific mitral valve representation that incorporates image‐derived material properties. We use our computational framework, along with 3D TEE images of the open and closed mitral valve, to estimate values for chordae rest lengths and leaflet material properties. These parameters are initialized using generic values and optimized to match the visualized deformation of mitral valve geometry between the open and closed states. Optimization is achieved by minimizing the summed Euclidean distances between the estimated and image‐derived closed mitral valve geometry. The spatially varying material parameters of the mitral leaflets are estimated using an extended Kalman filter to take advantage of the temporal information available from TEE. This semi‐automated and patient‐specific modeling framework was tested on 15 TEE image acquisitions from 14 patients. Simulated mitral valve closures yielded average errors (measured by point‐to‐point Euclidean distances) of 1.86 ± 1.24 mm. The estimated material parameters suggest that the anterior leaflet is stiffer than the posterior leaflet and that these properties vary between individuals, consistent with experimental observations described in the literature. HighlightsA semi‐automatic framework to build patient‐specific models of mitral valve from medical images under user guidance.Chordae rest length and material parameters are optimized to calibrate patient‐specific models to simulate mitral valve closure consistently with observations from TEE images.First study to our knowledge to estimate material parameters of mitral leaflets on humans from TEE images.Simulated mitral valve closure from 14 sets of images on 15 patients are compared to the ground truth estimated from TEE images with promising accuracy. Graphical abstract Figure. No caption available.

Flow Network Based Cardiac Motion Tracking Leveraging Learned Feature Matching

We present a novel cardiac motion tracking method where motion is modeled as flow through a network. The motion is subject to physiologically consistent constraints and solved using linear programming. An additional important contribution of our work is the use of a Siamese neural network to generate edge weights that guide the flow through the network. The Siamese network learns to detect and quantify similarity and dissimilarity between pairs of image patches corresponding to the graph nodes. Despite cardiac motion tracking being an inherently spatiotemporal problem, few methods reliably address it as such. Furthermore, many tracking algorithms depend on tedious feature engineering and metric refining. Our approach provides solutions to both of these problems. We benchmark our method against a few other approaches using a synthetic 4D echocardiography dataset and compare the performance of neural network based feature matching with other features. We also present preliminary results on data from 5 canine cases.

Multi-band confidence processing for two-pass speckle tracking

In myocardial strain imaging using speckle tracking, large interframe strains can cause significant peak hopping. An iterative, or multi-pass, process for speckle tracking, using displacement estimates from the previous pass as an initial guess for a subsequent pass, can improve displacement accuracy if initial guesses are close to the true correlation coefficient peak. Different peak-hopping patterns are observed when speckle tracking is performed on different frequency components of radiofrequency (RF) data. To reduce these artifacts, we propose a two-pass approach using a multi-band algorithm to assess the quality of the displacement guess at each pixel by comparing correlation results from multiple sub-bands. In this study, RF data were acquired from the anterior wall of the left-ventricle in an open-chest dog using a commercial 2-D phased array. RF images were filtered into five frequency bands. In the first-pass, 2-D phase-sensitive correlation-based tracking (search region 33 x 9 pixels (axial x lateral)) was applied to adjacent frames of the broadband image and each sub-band image to produce six sets of 2-D displacement estimates. Following the first pass, a confidence index was found for each pixel based on the number of sub-bands with matching displacement estimates, the magnitude of the correlation coefficient, and estimated strain values. Cubic interpolation was performed between pixels in the broadband image for which the confidence weight exceeded a threshold. Second-pass tracking using the resultant displacements was performed with a small search region (5 x 3 pixels (axial x lateral)) to reduce peak hopping. The two-pass approach with multi-band selection criteria reduced peak hopping compared to the single-pass method without sacrificing spatial resolution. With the two-pass method, the variance of the interframe displacement was reduced by over 96% axially and 98% laterally. The variance of axial and lateral displacements accumulated from end diastole to peak systole was reduced by 99%.