ICM Postdoctoral Fellow Dr. Siamak Ardekani has received a prestigious American Heart Association Fellowship from the Mid-Atlantic Affiliate Research Committee of the AHA. His research project is entitled “Algorithms for Detecting Changes in Heart Shape and Motion That Are Indicative of Disease State and Arrhythmia Risk” and he will be mentored jointly by Drs. Michael I Miller and Raimond L. Winslow. His research abstract is as follows:

Abstract:
Cardiac disease is often accompanied by ventricular remodeling, a process by which mechanical, neurohormonal, and genetic factors alter ventricular size, shape and function. A detailed knowledge of ventricular remodeling therefore has fundamental importance to our understanding of disease processes and to the development of new treatment strategies. The emerging discipline of Computational Anatomy (CA) is building a mathematical framework for describing anatomic variability and for performing statistical inference and hypothesis testing on disease-induced anatomic changes. In prior work, we used methods of CA to identify the nature of anatomic remodeling in the canine tachycardia pacing-induced model of heart failure by imaging and analyzing excised canine heart (sect. B & C). Results suggest the hypothesis that these analytical methods may be applied to in-vivo analysis of disease-induced changes of both heart structure and motion in a quantitative and objective fashion. If this hypothesis proves to be true, then a broad range of significant clinical applications would be possible. As but one example, an ongoing project in the D. W. Reynolds Cardiovascular Clinical Research Center at JHU is imaging heart shape and motion in a large cohort of patients with left ventricular (LV) dysfunction and who have received Implantable Cardioverter Defibrillator (ICD) therapy for primary prevention of sudden cardiac death (SCD). Quantitative cardiac shape and motion analysis may be useful in revealing those patients who are at highest risk for SCD and who therefore are in greatest need for ICD therapy.

Dilated cardiomyopathy (DCM) is one of the important causes of congestive heart failure (HF), the leading cause of hospitalization in older individuals in the United States. Regardless of type or cause of DCM, an initial insult that produces systolic dysfunction is followed by the initiation of processes designed to temporarily stabilize pump function. However, despite short-term benefits, these compensatory responses ultimately fail over the longer term. The majority of work has focused on characterizing the nature of end-stage DCM. Consequently, there is little information available regarding the time-evolution of anatomic and functional remodeling during DCM. Recently, Costandi et al examined the time course of global anatomical and functional changes that occur during transition to overt HF in a murine model (MLP knockout) of DCM. This study indicates that at the early stage of DCM, ventricular remodeling is mostly driven by myocardial hypertrophic growth. Shortly after this stage, the ventricles undergo a transient dilation followed by overt HF. While several factors, including a possible role of MLP abnormalities, have been attributed to human HF, the murine model of heart failure produces many of the clinical features of human DCM. The MLP deficient murine model of heart failure therefore provides us with a well-studied, controlled preparation with which we will develop, analyze and refine our computational algorithms for in-vivo characterization of changes in heart shape and motion. As our first major objective, we will test the hypothesis that methods of computational anatomy can be applied to in-vivo imaging data to detect functionally relevant changes in heart shape and motion in MLP knockout (MLPKO) murine model of DCM. To do this, we will develop new computational anatomy methods that, when applied to 3D time evolving heart imagery obtained from the MLPKO model of heart failure, enable us to detect statistically significant changes in heart shape and motion over time. As a second major objective, we will test the hypothesis that the analytical methods developed in the MLPKO study can be applied to in-vivo imaging data from a large cohort of patients who are at risk for SCD, who have received ICD placement and who are followed over time to determine those whose ICDs fire (high risk of SCD) and those whose ICDs do not fire (lower risk). We will attempt to find heart shape and motion changes that are predictive of those patients who are at high risk for SCD and thus who are best suited for ICD therapy.