Center for Integrative Biomedical Computing

Atrial fibrillation (AF) is an electrophysiological condition that represents an increasing problem in the aging populations of the world; AF doubles the risk of stroke and mortality and diminishes quality of life. The best current method to evaluate the progression of AF and monitor the success of interventions is via an invasive intra-cardiac catheter-based electrical mapping procedure. A noninvasive means to evaluate characteristics of AF prior to treatment and to track the effect of interventions over time would be extremely valuable, and magnetic resonance imaging (MRI) offers such an opportunity. Before MRI can achieve its potential, there are challenging technical problems to overcome, such as the high spatial resolution required to image the thin atrial wall and the temporal resolution and gating necessary to compensate for the distorting effects of respiratory and cardiac motion. The Comprehensive Arrhythmia Research and Management (CARMA) Center has become a world leader in the use of MRI in AF and has overcome many of the image acquisition hurdles to make MRI a standard component of AF patient management at our institution. These improvements in image acquisition have opened up significant opportunities and new questions for the understanding and clinical management of AF.

Atrial fibrillation is a growing problem in modern societies with an enormous impact on both short-term quality of life and long-term survival. Approximately 0.5% of people aged 50 to 59 have atrial fibrillation, and in populations aged 80 to 89, 9% are afflicted with AF. While many cases go untreated, AF is associated with an almost two-fold increase in the risk of mortality. AF patients experience a dramatically increased rate of stroke, from 1.5% for those aged 50 to 59 years to 23.5% for those aged between 80 and 89, a risk that, by contrast, decreases with age in the normal population. Treatment of AF represents a significant healthcare burden with the annual costs estimated at around 7 billion US dollars. Restoring and maintaining sinus rhythm remains one of the major goals in treating patients with AF. The inadequacies of drug-based treatments have long been the major motivation for finding a truly alternative approach to maintaining sinus rhythm and suppressing AF. Despite the fact that ablation, when successful, offers a complete and final cure, the success rate of ablation in maintaining regular sinus rhythm without the additional use of antiarrhythmic medications still remains at a mere 40-80%. Our research aims at increasing this rate.

Figure from Bauer al., MRI-based injury characterization immediately following ablation of AF. Creation of surface model for display and analysis of tissue enhancement states - (A.) The endocardial surface of the left atrium (blue), along with regions of the LA wall with different enhancement states (red >> hypoenhancement and green >> hyper-enhancement) were defined using Seg3D. (B.) A surface model generated from the segmentation of the endocardial surface. (C.) Addition of isovolumes generated from the tissue type segmentations (D.) Samples of the surrounding isovalues were recorded at discrete intervals (.2 mm) along normal vectors extending from the surface of the mesh 4 mm into the surrounding space. (E.) The maximum isovalue detected along normals mapped to the nodes on the model.
Dr. Marrouche is the founder and director of CARMA and he and his team continue to published prolifically on all aspects of imaging and AF. They have firmly established the "Utah AFib Score", an MRI-derived index of fibrosis in the atria that predicts outcomes using the interventional therapy of ablation (cite Oaks, Akoum). Atrial Ablation is a catheter-based treatment that uses some form of energy applied to the inner surface of the heart (the endocardium) to electrically isolate the pulmonary veins. The pulmonary veins flow into the left atrium and are considered the dominant site of the origin of AF. Upon completion of ablation, MRI imaging is also a useful tool to determine the degree of scar formation, and the CARMA group has developed the means to predict long-term ablation success, based on the extent to which pulmonary veins are completely encircled with scar that covers the complete atrial wall thickness.

The CIBC has been a parter with CARMA from the beginning and continues to provide a range of software support in tandem with progress in AF research. The main segmentation tool for processing MRI scans from patients and animal studies remains Seg3D and CorView, which supports for atrial images (see below for more details on CoreView). SCIRun has also been a valuable tool for CARMA research because of its great flexibility for all aspects of quantitative analysis, geometric manipulations, and visualizations. This close collaboration between CARMA and the CIBC continues, as we describe below.

Clinical Practice

The overarching goal of our research in atrial fibrillation (AF) is to improve all aspects of the diagnosis, treatment, and management of this condition by means of multimodal medical imaging and image processing. The current standard of treatment for atrial fibrillation includes imaging by fluoroscopy, echocardiography, and electroanatomical mapping, all of which augment the electrocardiographic findings from standard and Holter ECGs. In addition to the standard imaging modalities, MRI-based approaches are supporting rapid advances in all phases of the management of AF patients, especially with the use of contrast agents and novel MRI acquisition techniques. In this report, we briefly summarize some recent advances in our use of MRI for AF management with a special focus on how these findings affect the modeling and simulation of AF.

At the center of both pre-ablation evaluation of fibrosis and post-ablation visualization of scar is the use of advanced segmentation tools, encapsulated in the new version of Seg3D that has been recently released. When CARMA switched to Seg3D for segmentation of MRI images, they immediately achieved a more than 5-fold improvement in the time required to complete the quantitative analysis carried out as part of their AF staging process. Seg3D version 2 was developed with the needs of the CARMA group in mind, specifically, it has the ability to remove the generic and general purpose interface so that application-specific wrappers can provide a more detailed interface. Moreover, these wrappers can include any amount of additional functionality that is useful for a specific application. The interface to the core Seg3D functionality is well defined and well documented so that there is both reuse of the existing Seg3D functionality and a relatively straightforward way to access it. The result for the CARMA project was CorView, a program that encapsulates segmentation functionality from Seg3D but also data handling, domain specific interface, quantitative image analysis, and visualization capabilities that are required for use in the clinical setting. CorView was entirely a product of and funded by CARMA, and it contains proprietary elements, highlighting the value of the CIBC software license for open-source and proprietary projects.

Going forward, the impact of the proposed improvements to Seg3D will also be of great value to the CARMA projects. While the current Seg3D handles the present sizes of MRI scans on patients with atrial fibrillation, the plans of CARMA investigators to apply some of the same methods to the ventricles will require more capacity. Cine MRI provides a means to capture not only cardiac geometry but also its motion. By combining the current use of gadolinium enhancement, both early and late, with cine scans, CARMA members hope to provide new insights into the relationship between structure changes in heart tissue and ventricular arrhythmias. However, with multiple MRI modalities and time dependent scans that will be combined as part of the segmentation and quantification, the resulting data sets will become very large.

Perhaps even more relevant for future CARMA needs, are the plans to make use of complex algorithms for clinical applications. There is enormous pressure to automate the algorithms used for image analysis, both to increase efficiencies and to remove the inevitable biases that come with manual oversight. At present, the computational complexity of potential algorithms blocks the progress of such approaches. Investigators cannot carry out the necessary exploration of algorithm and parameter selection because of the computational overhead of each iteration they attempt. For the field of image-based evaluation and management of cardiac patients to continue to grow, CARMA, and other groups with similar goals, will need highly efficient, robust, and replicable techniques for segmentation.

Corview Software

Corview is a clinical image-analysis software developed at CARMA that implements late gadolinium enhancement MRI postprocessing workflows for AF management. Corview incorporates innovate image-processing algorithms and MRI volume visualization to produce detailed patient-specific models of heart structures, assess left-atrial tissue structural remodeling for AF staging, and assess the regional patterning and extent of post-radiofrequency ablation scarring. In support of other CARMA research efforts, Corview also serves as a platform for rapid prototyping and deployment of new cardiac image analysis. At CARMA, the unique embedding of dedicated software architects and image-processing scientists in the hospital environment has produced a streamlined tool for clinical work that has been designed, in part, by the clinicians themselves. Corview is currently in early release within the CARMA center and in daily use by CARMA MRI processing technicians. A beta release to selected research collaborators is planned for the summer of 2011.

The CorView application. CorView is a domain-specific application constructed by replacing the standard Seg3D interface with a set of capabilities specific to the needs of cardiac segmentation, image-based analysis, and visualization. The image shows an example of an interaction panel (left) for setting parameters; the Seg3D display window (middle) showing segmentations of the atrial wall (red) and enhanced, scarred regions within the wall (green); and the interactive volume visualization window (right) showing the complete segmentation of the three-dimensional scan of the left atrium. The lower panel contains a histogram of voxel intensity used to set thresholds for identifying enhancement in the images.

Animal Research: MRI guided ablation

CARMA also continues to pioneer the use of real- time MRI in interventional procedures, and CIBC software continues to be an essential component of this progress. One the major benefits of using MRI for real-time guidance is the ability to compose visualizations that include many different types of information and merge previously acquired data with updates. In the real- time MRI-guided ablation experiments carried out by CARMA, this process starts with an acquisition of high resolution magnetic resonance angiography (MRA) of the chambers of the heart, followed by a segmentation of the chambers using Seg3D and construction of a multidomain surface mesh using SCIRun. The resulting triangulated surfaces are then saved using a custom module in SCIRun to an XML format, passed back to the visualization software (IFE from Siemens Healthcare), and combined with the selected single images from the MRI scanner and the animation of the catheters, the position of which is extracted from microcoils built into the catheter body.

The latest generation of MRI compatible catheters also provides an electrical measurement capability so that it is possible to replicate the electroanatomical mapping features of the current clinical mapping systems for cardiac electrophysiology studies. CARMA investigators have used SCIRun to implement such a system and are currently comparing its accuracy to one of the commercial, non-MRI compatible systems. During the MRI guided ablation procedure, the operator can capture the local electrical signals at the tip of the catheter, along with its position in three dimensional space. By sequentially sampling the inner surface of the heart over time, it is possible to compose a map of parameters such as electrical signal amplitude and timing of the local activation of cardiac tissue. SCIRun gathers all this information, together with the previously generated surface meshes of the cardiac chambers, and applies interpolation and visualization techniques to create spatiotemporal maps of the parameters of interest. Such maps have a significant advantage over standard electroanatomical maps, as the underlying display surface is created from high resolution MRI images rather than sparse samplings of the catheter electrode.

Awards / Acknowledgement in 2010-2011

Recent focus on CARMA:


N. Akoum, M. Daccarett, C. McGann, N. Segerson, G. Vergara, S. Kuppahally, T. Badger, N. Burgon, T. Haslam, E. Kholmovski, R.S. MacLeod, and N.F. Marrouche. Atrial fibrosis helps select the appropriate patient and strategy in catheter ablation of atrial fibrillation: a DE-MRI guided approach. J. Cardiovasc. Electrophys. 22(1):16–22, 2011.

G.R. Vergara, S. Vijayakumar, E.G. Kholmovski, J.J. Blauer, M.A. Guttman, C. Gloschat, G. Payne, K. Vij, N.W. Akoum, M. Daccarett, C.J. McGann, R.S. Macleod, and N.F. Marrouche. Real-time magnetic resonance imaging-guided radiofrequency atrial ablation and visualization of lesion formation at 3 Tesla. Heart Rhythm 8(2):295-303, 2011.

M. Daccarett, T.J. Badger, N. Akoum, N.S. Burgon, C. Mahnkopf, G.R. Vergara, E.G. Khol- movski, C.J. McGann, D. Parker, J. Brachmann, R.S. Macleod RS, and Marrouche NF. Asso- ciation of left atrial fibrosis detected by delayed-enhancement magnetic resonance imaging and the risk of stroke in patients with atrial fibrillation. J Am Coll Cardiol 57(7):831-8, 2011.

M. Daccarett, C.J. McGann, N.W. Akoum, R.S. MacLeod, and N.F. Marrouche. MRI of the left atrium: predicting clinical outcomes in patients with atrial fibrillation. Expert Rev Cardiovasc Ther 9(1):105-11, 2011.

5. C. Mahnkopf, T.J. Badger, N.S. Burgon, M. Daccarett, T.S. Haslam, C.T. Badger, C.J. Mc- Gann, N. Akoum, E. Kholmovski, R.S. MacLeod, and N.F. Marrouche. Evaluation of the Left Atrial Substrate in Patients with Lone Atrial Fibrillation Using Delayed-Enhanced MRI: Implications for Disease Progression and Response to Catheter Ablation. Heart Rhythm. 7(10):1475-81, 2010.

T.J. Badger, M. Daccarett, N.W., Akoum, Y.A. Adjei-Poku, N.S. Burgon, T.S. Haslam, S. Kalvaitis, S. Kuppahally. G. Vergara L. McMullen, P.A. Anderson PA, E. Kholmovski, R.S. Macleod, and N.F. Marrouche. Evaluation of Left Atrial Lesions after Initial and Repeat Atrial Fibrillation Ablation: Lessons Learned from Delayed-Enhancement MRI in Repeat Ablation Procedures. Circ Arrhythm Electrophysiol 3(3):249-59, 2010.

S.S. Kuppahally. N. Akoum, N.S. Burgon, T.J. Badger, E.G. Kholmovski, S. Vijayakumar, S.N. Rao, J. Blauer, E.N. Fish, E.V. Dibella, R.S., Macleod, C. McGann, S.E. Litwin, and N.F. Marrouche. Left Atrial Strain and Strain Rate in Patients with Paroxysmal and Persistent Atrial Fibrillation: Relationship to Left Atrial Structural Remodeling Detected by Delayed Enhancement-MRI. Circ Cardiovasc Imaging 3(3):231–239, 2010.

N.M. Segerson, M. Daccarett, T.J. Badger, A. Shabaan, N. Akoum, E.N. Fish, S.N. Rao, N.S. Burgon, Y. Adjei-Poku, E. Kholmovski, S. Vijayakumar, E.V. Dibella, R.S. Macleod, and N.F. Marrouche NF. Magnetic Resonance Imaging-Confirmed Ablative Debulking of the Left Atrial Posterior Wall and Septum for Treatment of Persistent Atrial Fibrillation: Rationale and Initial Experience. J Cardiovasc Electrophysiol 21(2):126-132, 2010.