Shusil Dangi

Integrating Atlas and Graph Cut Methods for Left Ventricle Segmentation from Cardiac Cine MRI

Magnetic Resonance Imaging (MRI) has evolved as a clinical standard-of-care imaging modality for cardiac morphology, function assessment, and guidance of cardiac interventions. All these applications rely on accurate extraction of the myocardial tissue and blood pool from the imaging data. Here we propose a framework for left ventricle (LV) segmentation from cardiac cine MRI. First, we segment the LV blood pool using iterative graph cuts, and subsequently use this information to segment the myocardium. We formulate the segmentation procedure as an energy minimization problem in a graph subject to the shape prior obtained by label propagation from an average atlas using affine registration. The proposed framework has been validated on 30 patient cardiac cine MRI datasets available through the STACOM LV segmentation challenge and yielded fast, robust, and accurate segmentation results.

Using Atlas Prior with Graph Cut Methods for Right Ventricle Segmentation from Cardiac MRI

Right ventricle segmentation helps quantify many functional parameters of the heart and construct anatomical models for intervention planning. Here we propose a fast and accurate graph cut segmentation algorithm to extract the right ventricle from cine cardiac MRI sequences. A shape prior obtained by propagating the right ventricle label from an average atlas via affine registration is incorporated into the graph energy. The optimal segmentation obtained from the graph cut is iteratively refined to produce the final right ventricle blood pool segmentation. We evaluate our segmentation results against gold-standard expert manual segmentation of 16 cine MRI datasets available through the MICCAI 2012 Cardiac MR Right Ventricle Segmentation Challenge. Our method achieved an average Dice Index 0.83, a Jaccard Index 0.75, Mean absolute distance of 5.50 mm, and a Hausdorff distance of 10.00 mm.

Automatic LV Feature Detection and Blood-Pool Tracking from Multi-plane TEE Time Series

Multi-plane, 2D TEE images constitute the clinical standard of care for assessment of left ventricle function, as well as for guiding various minimally invasive procedure that rely on intra-operative imaging for real-time visualization. We propose a framework that enables automatic, rapid and accurate endocardial left ventricle feature identification and blood-pool segmentation using a combination of image filtering, graph cut, non-rigid registration-based motion extraction, and 3D LV geometry reconstruction techniques applied to the TEE image series. We evaluate our proposed framework using several retrospective patient tri-plane TEE image sequences and demonstrate comparable results to those achieved by expert manual segmentation using clinical software.

Cine cardiac MRI slice misalignment correction towards full 3D left ventricle segmentation.

Accurate segmentation of the left ventricle (LV) blood-pool and myocardium is required to compute cardiac function assessment parameters or generate personalized cardiac models for pre-operative planning of minimally invasive therapy. Cardiac Cine Magnetic Resonance Imaging (MRI) is the preferred modality for high resolution cardiac imaging thanks to its capability of imaging the heart throughout the cardiac cycle, while providing tissue contrast superior to other imaging modalities without ionizing radiation. However, there exists an inevitable misalignment between the slices in cine MRI due to the 2D + time acquisition, rendering 3D segmentation methods ineffective. A large part of published work on cardiac MR image segmentation focuses on 2D segmentation methods that yield good results in mid-slices, however with less accurate results for the apical and basal slices. Here, we propose an algorithm to correct for the slice misalignment using a Convolutional Neural Network (CNN)-based regression method, and then perform a 3D graph-cut based segmentation of the LV using atlas shape prior. Our algorithm is able to reduce the median slice misalignment error from 3.13 to 2.07 pixels, and obtain the blood-pool segmentation with an accuracy characterized by a 0.904 mean dice overlap and 0.56 mm mean surface distance with respect to the gold-standard blood-pool segmentation for 9 test cine MR datasets.

3D printing for orthopedic applications: from high resolution cone beam CT images to life size physical models

With increasing resolution in image acquisition, the project explores capabilities of printing toward faithfully reflecting detail and features depicted in medical images. To improve safety and efficiency of orthopedic surgery and spatial conceptualization in training and education, this project focused on generating virtual models of orthopedic anatomy from clinical quality computed tomography (CT) image datasets and manufacturing life-size physical models of the anatomy using 3D printing tools. Beginning with raw micro CT data, several image segmentation techniques including thresholding, edge recognition, and region-growing algorithms available in packages such as ITK-SNAP, MITK, or Mimics, were utilized to separate bone from surrounding soft tissue. After converting the resulting data to a standard 3D printing format, stereolithography (STL), the STL file was edited using Meshlab, Netfabb, and Meshmixer. The editing process was necessary to ensure a fully connected surface (no loose elements), positive volume with manifold geometry (geometry possible in the 3D physical world), and a single, closed shell. The resulting surface was then imported into a “slicing” software to scale and orient for printing on a Flashforge Creator Pro. In printing, relationships between orientation, print bed volume, model quality, material use and cost, and print time were considered. We generated anatomical models of the hand, elbow, knee, ankle, and foot from both low-dose high-resolution cone-beam CT images acquired using the soon to be released scanner developed by Carestream, as well as scaled models of the skeletal anatomy of the arm and leg, together with life-size models of the hand and foot.

Integrating atlas and graph cut methods for right ventricle blood-pool segmentation from cardiac cine MRI

Segmentation of right ventricle from cardiac MRI images can be used to build pre-operative anatomical heart models to precisely identify regions of interest during minimally invasive therapy. Furthermore, many functional parameters of right heart such as right ventricular volume, ejection fraction, myocardial mass and thickness can also be assessed from the segmented images. To obtain an accurate and computationally efficient segmentation of right ventricle from cardiac cine MRI, we propose a segmentation algorithm formulated as an energy minimization problem in a graph. Shape prior obtained by propagating label from an average atlas using affine registration is incorporated into the graph framework to overcome problems in ill-defined image regions. The optimal segmentation corresponding to the labeling with minimum energy configuration of the graph is obtained via graph-cuts and is iteratively refined to produce the final right ventricle blood pool segmentation. We quantitatively compare the segmentation results obtained from our algorithm to the provided gold-standard expert manual segmentation for 16 cine-MRI datasets available through the MICCAI 2012 Cardiac MR Right Ventricle Segmentation Challenge according to several similarity metrics, including Dice coefficient, Jaccard coefficient, Hausdorff distance, and Mean absolute distance error.

Robust Head CT Image Registration Pipeline for Craniosynostosis Skull Correction Surgery.

Craniosynostosis is a congenital malformation of the infant skull typically treated via corrective surgery. To accurately quantify the extent of deformation and identify the optimal correction strategy, the patient-specific skull model extracted from a pre-surgical computed tomography (CT) image needs to be registered to an atlas of head CT images representative of normal subjects. Here, the authors present a robust multi-stage, multi-resolution registration pipeline to map a patient-specific CT image to the atlas space of normal CT images. The proposed registration pipeline first performs an initial optimisation at very low resolution to yield a good initial alignment that is subsequently refined at high resolution. They demonstrate the robustness of the proposed method by evaluating its performance on 560 head CT images of 320 normal subjects and 240 craniosynostosis patients and show a success rate of 92.8 and 94.2%, respectively. Their method achieved a mean surface-to-surface distance between the patient and template skull of <2.5 mm in the targeted skull region across both the normal subjects and patients.

Assessing Cardiac Tissue Function via Action Potential Wave Imaging Using Cardiac Displacement Data

The ability to visualize action potentials deep within the walls of the heart has important applications. It enables the identification of regions of electrically and mechanically compromised tissue that can mark the location(s) of infarcted and ischemic myocardial tissue, and also permits the visualization of normal and abnormal action potential wave propagation patterns for use in both clinical and cardiac research settings. Recently, we have been investigating the possibility of using 4-D mechanical deformation data, obtained either from MRI or ultrasound images, to reverse-calculate these action potential patterns [2, 4, 5]. This idea has also been studied by Konofagou et al. [6], who used mixed time and space second derivatives in the displacement fields to identify the location of action potentials. While this mixed-derivative method should be effective for spatially one-dimensional action potentials, it is less effective when propagation of the waves is fundamentally three-dimensional.

A comparison study of atlas-based 3D cardiac MRI segmentation: global versus global and local transformations

Magnetic Resonance Imaging (MRI) is a standard-of-care imaging modality for cardiac function assessment and guidance of cardiac interventions thanks to its high image quality and lack of exposure to ionizing radiation. Cardiac health parameters such as left ventricular volume, ejection fraction, myocardial mass, thickness, and strain can be assessed by segmenting the heart from cardiac MRI images. Furthermore, the segmented pre-operative anatomical heart models can be used to precisely identify regions of interest to be treated during minimally invasive therapy. Hence, the use of accurate and computationally efficient segmentation techniques is critical, especially for intra-procedural guidance applications that rely on the peri-operative segmentation of subject-specific datasets without delaying the procedure workflow. Atlas-based segmentation incorporates prior knowledge of the anatomy of interest from expertly annotated image datasets. Typically, the ground truth atlas label is propagated to a test image using a combination of global and local registration. The high computational cost of non-rigid registration motivated us to obtain an initial segmentation using global transformations based on an atlas of the left ventricle from a population of patient MRI images and refine it using well developed technique based on graph cuts. Here we quantitatively compare the segmentations obtained from the global and global plus local atlases and refined using graph cut-based techniques with the expert segmentations according to several similarity metrics, including Dice correlation coefficient, Jaccard coefficient, Hausdorff distance, and Mean absolute distance error.