Catálogo de publicaciones - libros

Experimental and inverse approaches have been applied in studying the contributions of different parts of the myocardium to the ECG measurements. Also optimal electrode locations for different clinical purposes have been studied by applying body surface maps. It is valuable to know where the measured ECG is actually generated. Thus the measurements can be designed to be most optimal to measure certain myocardial sources. Here we assess the contributions of 12 left ventricular segments to the potentials of 117 surface leads. The study is based on the numerical lead field analysis combined with the cardiac activation modeling. We analyzed the contributions of the signals generated by different segments to the total signal generated by the left ventricle. It was found that anterior segments have high contributions to the leads on the lower left thorax and inferior segments on the leads on the lower left back. These results were expected based on the previous clinical studies.

Non-invasive electrocardiographic (ECG) techniques for assessing the electrical activity of selected regions within the cardiac muscle can benefit from suitable positioning of surface electrodes. This positioning is usually guided heuristically and complemented by clinical and experimental studies, but there is a lack of general methods to characterize quantitatively the ability of a given electrode configuration to focus on selected regions of the heart. In this study we explore an approach to the characterization of the resolution of surface ECG systems based on the concept of Resolution Mass (RM). By integrating bioelectric signal modeling and numerical methods, we explore, in an application example, the location and size of the RM for a multielectrode ECG system. The concept of RM combined with bioelectric signal modeling and numerical methods constitutes a powerful tool to investigate the resolution properties of surface ECG systems.

We present an experimental preparation that allows simultaneous, high resolution mapping of endocardial, epicardial and torso tank surfaces from an isolated canine heart. The preparation additionally permits control over blood flow rate through a targeted region of the cardiac tissue, the heart rate and the mechanical load on the left ventricle. Adjustment of heart rate and left-ventricular load through a fluid-filled balloon inserted into the left ventricle allow control of metabolic demand. Cardiac potential measurements occur by means of flexible sock-mounted electrode arrays applied to the epicardium and the intraventricular balloon. A preliminary study using this preparation suggests the existence of a heterogenous response of the myocardium to ischemia. Such an experimental model is a useful testbed for many studies, including validation of forward and inverse solutions.

Background: Myocardial ischemia in the left ventricular (LV) myocardium introduces non-uniform and pathological strain patterns. Total passive segments lengthen when pressure increases during early systole and shorten when pressure drops at the end of ejection. Moderately ischemic segments typically lengthen during isovolumic contraction, start shortening during ejection and continue shortening, usually at an increased rate, after aortic valve closure. Aim: The aim of this study was to investigate possible mechanisms for the characteristic strain patterns in moderately ischemic regions using a simulation model of the LV wall. Methods: A thick-walled truncated ellipsoidal finite element model was used to represent the LV geometry. The model included mathematical descriptions of fiber orientation, passive elastic properties, and actively generated fiber stress. A severely ischemic region and a moderately ischemic border zone were incorporated in the model. The severely ischemic region was made stiffer and generated no active fiber stress during systole. The border zone was made slightly stiffer, active fiber stress was reduced and generated at a slower rate while the relaxation rate was slower than in the normal regions. The cardiac cycle was simulated by applying physiological pressure-volume boundary conditions. Results: The strain pattern in the severely ischemic region resembled the pressure curve with lengthening during pressure rise and shortening during pressure decrease, while the border zone started shortening after an initial early systolic lengthening and continued shortening during isovolumic relaxation at an increased rate. Conclusion: The characteristic moderately ischemic strain pattern may be caused by slower mechanical activation and relaxation rates.

Long term responses of the heart to e.g. infarction or surgical intervention are related to response of the tissue to changes in the mechanical environment. The tissue response is likely to involve (local) change of mass. Implementation of the associated inhomogeneous change in volume for a complex geometry is cumbersome. In the present study, we propose a computational framework for finite volumetric growth. The local stimulus for growth is determined from a simulation of beat to beat cardiac mechanics, assuming the tissue to be incompressible. The related local volumetric growth is translated in a global change of cardiac shape through a simulation of long term cardiac mechanics, assuming the tissue to be compressible. We illustrate the model by simulating growth in response to a deviation of end-diastolic sarcomeric strain from a set optimal value assumed to be preferred by the tissue. Inhomogeneity in the stimulus was reduced after inhomogeneous growth of up to 25%. The transmural redistribution of mass due to growth was found to alter an initially unphysiological linear transmural course in myofiber orientations to a more physiological course. We conclude that the model enables simulation of locally inhomogeneous growth in a realistic left ventricular geometry.

Clinical trials evaluating cardiac resynchronization therapy (CRT) have demonstrated that 30% of patients with heart failure and wide QRS do not respond to CRT (especially patients with myocardial infarcts). We have developed 3D numerical models of failing hearts, with and without chronic infarcts in different regions of the left ventricle. The hearts were coupled to a closed circulation, and the model included effects of the pericardium. The hearts were either paced at the right ventricular apex (RVA) or left ventricular free wall (LVFW). In normal and failing hearts, LV pump function was moderately better for LVFW pacing. In the normal heart model, heterogeneity of ejection strain was similar for RVA and LVFW pacing. However, in the failing heart model, LVFW pacing was associated with 44% less heterogeneity of ejection strain. This may be an important factor in the remodeling process associated with pacing.

We present a novel strategy to perform estimation for a mechanical system defined to feature the same essential characteristics as a heart model, using measurements of a type that is available in medical imaging. We adopt a sequential approach, and the joint state-parameter estimation procedure is constructed based on a robust and effective state estimator inspired from collocated feedback control. The convergence of the resulting joint estimator can be mathematically established, and we demonstrate its effectiveness by identifying localized contractility and stiffness parameters in a test problem representative of cardiac behavior and using synthetic –albeit realistic –measurements.

Placement of Implantable Cardiac Defibrillator (ICD) leads in children and some adults is challenging due to anatomical factors. As a result, novel ad hoc non-transvenous implant techniques have been employed clinically. We describe an open-source subject-specific, image-based finite element modeling software environment whose long term goal is determining optimal electrode placement in special populations of adults and children Segmented image-based finite element models of two children and one adult were created from CT scans and appropriate tissue conductivities were assigned. The environment incorporates an interactive electrode placement system with a library of clinically-based, user-configurable electrodes. Finite element models are created from the electrode poses within the torsos and the resulting electric fields, current, and voltages computed and visualized.

Rhythm disturbances and mechanical function suppression proper to the acute heart failure in the case of cardiomyocyte calcium overload are simulated in a mathematical model of cardiomyocyte electromechanical activity. Particular attention is paid to the overload caused by diminished activity of the Na^ + - K^ +  pump. It is shown in the framework of the model that myocardium mechanics may promote arrhythmias in these conditions. In particular, cooperative influence of the attached crossbridges on the calciumtroponin kinetics is shown to contribute to the initiation of spontaneous action potentials. Numerical experiments showed that the recovery of the normal Na^ + - K^ +  pump activity during the heart failure attack did not always led to the normal electromechanical function recovery in the failed cardiomyocyte. Alternative approaches were suggested in the model and compared to each other for recovery of the myocardium electrical and mechanical performance in the simulated case of the acute heart failure.

We recently introduced a continuous state space parametric model of spatio-temporal transformations and an algorithm, based on Kalman filtering, to represent motion in an image sequence describing a periodic phenomena. One advantage of this method is to simultaneously take into account all the sequence frames to robustly estimate the parameters of a unique spatial and periodic-temporal model. However, in 3D+time, a large number of parameters is required. In this paper, we propose a criterion based on motion energy to locally adapt the trajectory model and thus the temporal complexity of the model. The influence of the model order is illustrated on true 2D+time Magnetic Resonance Images (MRI) of the heart in order to motivate the proposed adaptative criteria. Quantitative results of the proposed adapted spatio-temporal motion model are given on synthetic 2D+time MRI sequences. Preliminary experiments show a significant impact notably regarding the parameter saving while preserving the accuracy of the motion estimates.