Nripesh Parajuli

Statistical Shape Modeling of the Left Ventricle: Myocardial Infarct Classification Challenge

Statistical shape modeling is a powerful tool for visualizing and quantifying geometric and functional patterns of the heart. After myocardial infarction (MI), the left ventricle typically remodels in response to physiological challenges. Several methods have been proposed in the literature to describe statistical shape changes. Which method best characterizes the left ventricular remodeling after MI is an open research question. A better descriptor of remodeling is expected to provide a more accurate evaluation of disease status in MI patients. We therefore designed a challenge to test shape characterization in MI given a set of three-dimensional left ventricular surface points. The training set comprised 100 MI patients, and 100 asymptomatic volunteers (AV). The challenge was initiated in 2015 at the Statistical Atlases and Computational Models of the Heart workshop, in conjunction with the MICCAI conference. The training set with labels was provided to participants, who were asked to submit the likelihood of MI from a different (validation) set of 200 cases (100 AV and 100 MI). Sensitivity, specificity, accuracy, and area under the receiver operating characteristic curve were used as the outcome measures. The goals of this challenge were to 1) establish a common dataset for evaluating statistical shape modeling algorithms in MI, and 2) test whether statistical shape modeling provides additional information characterizing MI patients over standard clinical measures. Eleven groups with a wide variety of classification and feature extraction approaches participated in this challenge. All methods achieved excellent classification results with accuracy ranges from 0.83 to 0.98. The areas under the receiver operating characteristic curves were all above 0.90. Four methods showed significantly higher performance than standard clinical measures. The dataset and software for evaluation are available from the Cardiac Atlas Project website.11 http://www.cardiacatlas.org.

Sparsity and Biomechanics Inspired Integration of Shape and Speckle Tracking for Cardiac Deformation Analysis

Cardiac motion analysis, particularly of the left ventricle (LV), can provide valuable information regarding the functional state of the heart. We propose a strategy of combining shape tracking and speckle tracking based displacements to calculate the dense deformation field of the myocardium. We introduce the use and effects of l1 regularization, which induces sparsity, in our integration method. We also introduce regularization to make the dense fields more adhering to cardiac biomechanics. Finally, we motivate the necessity of temporal coherence in the dense fields and demonstrate a way of doing so. We test our method on ultrasound (US) images acquired from six open-chested canine hearts. Baseline and post-occlusion strain results are presented for an animal, where we were able to detect significant change in the ischemic region. Six sets of strain results were also compared to strains obtained from tagged magnetic resonance (MR) data. Median correlation (with MR-tagging) coefficients of 0.73 and 0.82 were obtained for radial and circumferential strains respectively.

Integrated Dynamic Shape Tracking and RF Speckle Tracking for Cardiac Motion Analysis

We present a novel dynamic shape tracking (DST) method that solves for Lagrangian motion trajectories originating at the left ventricle (LV) boundary surfaces using a graphical structure and Dijkstra’s shortest path algorithm.

Left Ventricle Classification Using Active Shape Model and Support Vector Machine

We present the methods and results of building an Active Shape Model ASM of the left ventricle LV and using a Support Vector Machine SVM learning model to classify normal and infarcted LVs provided in the STACOM 2015 Statistical Shape Analysis Challenge dataset. First, all LVs are rigidly registered to a reference LV. In this way, the entire dataset is aligned to the same reference frame. Then, the shape model is obtained by Principal Component Analysis PCA decomposition of the aligned LVs. This allows us to capture the principal modes of variation in the LV shapes and reduce the dimensionality of our data. Next, we train an SVM learning model on our data. To test the performance of the model before using it on the unlabeled test set, we test our method by partitioning the dataset with labels into a training set and a test set of equal sizes. We train the model only using the training set and predict the labels of the test set we created whose labels were known but not used during training. We repeated this for 100 different random partitions and achieved 94i?ź% prediction accuracy with 94i?ź% sensitivity and 93i?ź% specificity on the test set.

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.

Learning-Based Spatiotemporal Regularization and Integration of Tracking Methods for Regional 4D Cardiac Deformation Analysis

Dense cardiac motion tracking and deformation analysis from echocardiography is important for detection and localization of myocardial dysfunction. However, tracking methods are often unreliable due to inherent ultrasound imaging properties. In this work, we propose a new data-driven spatiotemporal regularization strategy. We generate 4D Lagrangian displacement patches from different input sources as training data and learn the regularization procedure via a multi-layered perceptron (MLP) network. The learned regularization procedure is applied to initial noisy tracking results. We further propose a framework for integrating tracking methods to produce better overall estimations. We demonstrate the utility of this approach on block-matching, surface tracking, and free-form deformation-based methods. Finally, we quantitatively and qualitatively evaluate our performance on both tracking and strain accuracy using both synthetic and in vivo data.

Dictionary learning-based spatiotemporal regularization for 3D dense speckle tracking

Speckle tracking is a common method for non-rigid tissue motion analysis in 3D echocardiography, where unique texture patterns are tracked through the cardiac cycle. However, poor tracking often occurs due to inherent ultrasound issues, such as image artifacts and speckle decorrelation; thus regularization is required. Various methods, such as optical flow, elastic registration, and block matching techniques have been proposed to track speckle motion. Such methods typically apply spatial and temporal regularization in a separate manner. In this paper, we propose a joint spatiotemporal regularization method based on an adaptive dictionary representation of the dense 3D+time Lagrangian motion field. Sparse dictionaries have good signal adaptive and noise-reduction properties; however, they are prone to quantization errors. Our method takes advantage of the desirable noise suppression, while avoiding the undesirable quantization error. The idea is to enforce regularization only on the poorly tracked trajectories. Specifically, our method 1.) builds data-driven 4-dimensional dictionary of Lagrangian displacements using sparse learning, 2.) automatically identifies poorly tracked trajectories (outliers) based on sparse reconstruction errors, and 3.) performs sparse reconstruction of the outliers only. Our approach can be applied on dense Lagrangian motion fields calculated by any method. We demonstrate the effectiveness of our approach on a baseline block matching speckle tracking and evaluate performance of the proposed algorithm using tracking and strain accuracy analysis.

Automatic extraction of disease-specific features from Doppler images

An automatic extraction of disease-specific features from Doppler images to help diagnose valvular diseases is provided. The method includes the steps of obtaining a raw Doppler image from a series of images of an echocardiogram, isolating a region of interest from the raw Doppler image, the region of interest including a Doppler image and an ECG signal, and depicting at least one heart cycle, determining a velocity envelope of the Doppler image in the region of interest, extracting the ECG signal to synchronize the ECG signal with the Doppler age over the at least one heart cycle, within the region of interest, calculating a value of a clinical feature based on the extracted ECG signal synchronized with the velocity envelope, and comparing the value of the clinical feature with clinical guidelines associated with the clinical feature to determine a diagnosis of a disease.