Heart failure is the clinical syndrome accompanying the inability of the heart to maintain a cardiac output required to meet the metabolic requirements and accommodate venous return, and is one of the leading causes of mortality in United States. Accurate imaging of the heart and its failure is important for successful patient management and treatment. Multiple cardiac imaging modalities provide complementary information about the heart – LV function and wall motion, anatomy, myocardial viability and ischemia. In many instances, it is necessary for a patient to undergo multiple imaging sessions to obtain diagnostic clinical information with confidence. It would be beneficial to the individual and the health care system if a single imaging modality could yield reliable clinical information about the heart, leading to a reduced cost, anxiety and an increased diagnostic confidence. This thesis proposes methods that would make cardiac MRI perform an improved assessment of LV function, wall motion, and viability, such that cardiac MRI is taken one step closer to being a single stop solution for imaging of heart. Conventional cardiac MR imaging is performed at a temporal resolution of around 40 ms per cardiac phase. While the global left ventricular (LV) function can be reliably established at this temporal resolution, functional metrics characterizing transient function like peak filling and ejection rates are not accurately assessed. A high temporal resolution is necessary to characterize such transient LV function and wall motion mechanics. This thesis proposes methods to acquire cine-images of the heart at a higher temporal resolution (~ 6 ms) and algorithms to acquire the LV volume across all cardiac phases that would yield functional metrics characterizing LV function and wall motion mechanics. The validation of these algorithms was performed on human subjects. Cardiac MR imaging is the current gold standard of myocardial viability imaging, in which scarred regions of the heart following myocardial infarction are visualized. However viability imaging faces image quality challenges in patients with severe arrhythmias and in cases where a higher spatial resolution, and hence a longer acquisition time, is desired. This thesis also proposes an arrhythmia insensitive inversion recovery (AIIR) algorithm that would significantly reduce artifacts that degrade image quality, thereby extending viability imaging to higher spatial resolution and in patients with severe arrhythmia. Simulations, experimental validation on phantoms and clinical verification on patients are performed. Results from high temporal resolution imaging reveal that obtaining cine cardiac MR images at a temporal resolution of 6 ms per cardiac phase is feasible. Appropriate validated algorithms yield LV time-volume curve from which LV functional metrics are reliably extracted. A dependence on temporal resolution is revealed, and a temporal resolution cut-off of 12 ms is proposed to reliably capture the temporal dynamics of the LV. Also, results from cardiac viability imaging show that the AIIR algorithm performs significantly better than conventional imaging methods in both phantoms and human subjects, as shown by the blinded expert scores, leading to a better image quality. In conclusion, this thesis proposes and implements methods that help cardiac MRI yield 1) a better function and wall motion assessment of the heart through high temporal resolution imaging and 2) a better assessment of myocardial viability through the AIIR algorithm.

When excitable cardiac tissue is electrically paced at a sufficiently high rate, the duration of excitation can alternate from beat to beat despite a constant stimulation period. This rhythm, known as alternans, has been identified as an early stage in a sequence of increasingly complex instabilities leading to the lethal arrhythmia ventricular fibrillation (VF). This connection served as as a motivation for research into the control of alternans as a strategy to prevent VF. Control methods that do not use a model of the dynamics have been used for the suppression of alternans. However, these methods possess limitations. In this thesis we study theoretically model-based control techniques with the goal of developing protocols that would overcome the shortcomings of non model-based approaches. We consider one dimensional tissue in two different geometrical configurations: a ring and a fiber with free ends (open fiber). We apply standard control methods for linear time invariant systems to a stroboscopic map of the linearized dynamics around the normal rhythm. We found that, in the ring geometry, model-based control is able to suppress alternans faster and with lower current, thereby reducing the risk of tissue damage, compared with non-model-based control. In the open fiber, model-based control is able to suppress alternans for longer fibers and higher pacing frequencies in comparison with non-model-based control. The methodology presented here can be extended to two- and three-dimensional tissue, and could eventually lead to the suppression of alternans on the entire ventricles.

Perseverative cognition is the repetitive cognitive representation of a stressor, which includes the concepts of worry and rumination. These thoughts delay post-stress cardiovascular recovery, which may lead to an increased risk for cardiovascular disease. This may be due to the negative emotional content of perseverative cognition or because it involves cognitive effort. The aim of this study was to identify the unique influences of negative emotional content and cognitive effort during recovery. Participants (N = 120) were given a demanding task purportedly as a measure of intelligence and then given false negative feedback. Immediately following, participants engaged in one of four recovery instruction conditions: think about task performance, perform a cognitive load task, watch a distracting video, or remain quietly seated. EKG, impedance cardiography, and blood pressure were recorded throughout. Perseverative cognition and cognitive load both resulted in significantly less heart rate recovery compared to the distracting video. Higher test motivation and anxiety were related to more blunted reactivity and delayed recovery of respiratory sinus arrhythmia. Reduced recovery during perseverative cognition and cognitive effort indicate that the cognitive load produced by perseveration is the pernicious component that explains its link to increased risk for cardiovascular disease. Further, the relationship between motivation/anxiety and blunted reactivity and recovery suggest effort may be important in the link between perseverative cognition and cardiovascular disease.

In the heart, electrical signaling is responsible for its rhythmicity and is necessary to initiate muscle contraction. The net electrical activity in a cardiac myocyte during a contraction cycle is observed as the action potential (AP), which describes a change in membrane potential as a function of time. In ventricular cardiac myocytes, voltage-gated sodium channels (Nav) and voltage-gated potassium channels (Kv) play antagonistic roles in shaping the AP with the former initiating membrane depolarization and the latter repolarizing it. Functional changes in the primary cardiac Nav isoform, Nav 1.5, or any one of the many Kv isoforms expressed in the ventricle, as evidenced by those characterized in various congenital and/or acquired etiologies, can lead to severe cardiac pathologies. Nav and Kv are large transmembrane proteins that can be extensively post-translationally modified through processes that include glycosylation. The sequential glycosylation process typically ends with negatively charged sialic acid residues added through trans-Golgi sialyltransferase activity. Sialyltransferases belong to a much larger group of glycogene products that number in the hundreds and are responsible for creating a complex and variable glycan profile (glycome) unique to different cell types and tissues. Sialic acids impact Nav and Kv function likely by contributing to the extracellular surface potential and thereby causing channels to gate following smaller depolarizations. Additionally, developmentally regulated sialylation contributes to cardiac myocyte excitability in the neonatal mouse atria. However, little is understood concerning how the glycosylation machinery (glycogene products) influences cell and tissue electrical signaling. The sialytransferase Β-galactoside α-2,3-sialyltransferase 4 (ST3Gal4) adds sialic acids to galactose residues of N- and O-linked glycans through α-2,3-linkgages. ST3Gal4 is uniformly expressed throughout the chambers and developmental stages of the heart and therefore is likely a useful target to question whether and how glycosylation impacts these events. Additionally, diseases of glycosylation often cause symptoms that are consistent with changes in excitability that include arrhythmias and seizures. Congenital disorders of glycosylation lead to variably reduced glycoprotein and glycolipid glycosylation. However, because sialic acids are typically the terminal residues added to glycan structures, disease-related reduced glycosylation often leads to fewer sialic acids being attached. In addition, Chagas disease, which results in pathological changes in cardiac electrical function, may reduce sialic acids directly. Because of this, the ST3Gal4-/- strain was also used to investigate the role of glycosylation in the pathological cardiac electrical remodeling often associated with these diseases. The methodologies included cellular, tissue and whole-animal electrophysiology as well as biochemical assays. The data indicate that deletion of ST3Gal4 significantly affects Nav sialylation and gating with no change in maximum current density or protein expression. ST3Gal4 deletion also depolarizes the activation gating of both voltage-dependent kinetic components of repolarization found in the mouse ventricle: Ito and IKslow; however unlike the effect on INa, ST3Gal4 gene deletion causes a reduction in the peak IK density. Protein expression of the putative Kv isoforms responsible for Ito and IKslow was variably affected by ST3Gal4 gene deletion with Kv1.5 and Kv4.2 demonstrating no differences in protein densities. Contrastingly, a small but significant reduction in Kv2.1 protein from ST3Gal4-/- ventricular tissue was observed. In addition to effects on Nav and Kv activity, ST3Gal4 expression is necessary for normal cellular electrical signaling as demonstrated by a reduction in cellular refractory period and alterations in AP waveforms that include a slowing of cellular conduction and an extension of AP duration in ventricular myocytes from ST3Gal4-/- mice. Concurrent with aberrant excitability at the cellular level, the ST3Gal4-/- left ventricular epicardium demonstrated a reduced refractory period and was more susceptible to arrhythmias as observed through optical mapping studies. Additionally, ECGs of ambulatory ST3Gal4-/- mice demonstrated that deletion of the gene causes modest aberrant conduction under basal conditions and, in preliminary studies, appears to increase susceptibility to arrhythmias following a cardiac challenge, in the form of a low dosage of the Β-adrenergic agonist isoproterenol, suggesting a reduction in repolarization reserve in ST3Gal4 hearts. Based on the data reported here, it is apparent that relatively minor perturbations in the cardiac glycome cause significant changes in cardiac electrical signaling. These data highlight the role of glycosylation in normal physiology and underscore it as an important mediator in diseases where it may be altered.

published_or_final_version / abstract / Electrical and Electronic Engineering / Master / Master of Philosophy

Ultrasound Current Source Density Imaging (UCSDI) is a noninvasive modality for mapping electrical activities in the body (brain and heart) in 4-dimensions (space + time). Conventional cardiac mapping technologies for guiding the radiofrequency ablation procedure for treatment of cardiac arrhythmias have certain limitations. UCSDI can potentially overcome these limitations and enhance the electrophysiology mapping of the heart. UCSDI exploits the acoustoelectric (AE) effect, an interaction between ultrasound pressure and electrical resistivity. When an ultrasound beam intersects a current path in a material, the local resistivity of the material is modulated by the ultrasonic pressure, and a change in voltage signal can be detected based on Ohm's Law. The degree of modulation is determined by the AE interaction constant K. K is a fundamental property of any type of material, and directly affects the amplitude of the AE signal detected in UCSDI. UCSDI requires detecting a small AE signal associated with electrocardiogram. So sensitivity becomes a major challenge for transferring UCSDI to the clinic. This dissertation will determine the limits of sensitivity and resolution for UCSDI, balancing the tradeoff between them by finding the optimal parameters for electrical cardiac mapping, and finally test the optimized system in a realistic setting. This work begins by describing a technique for measuring K, the AE interaction constant, in ionic solution and biological tissue, and reporting the value of K in excised rabbit cardiac tissue for the first time. K was found to be strongly dependent on concentration for the divalent salt CuSO₄, but not for the monovalent salt NaCl, consistent with their different chemical properties. In the rabbit heart tissue, K was determined to be 0.041 ± 0.012 %/MPa, similar to the measurement of K in physiologic saline: 0.034 ± 0.003 %/MPa. Next, this dissertation investigates the sensitivity limit of UCSDI by quantifying the relation between the recording electrode distance and the measured AE signal amplitude in gel phantoms and excised porcine heart tissue using a clinical intracardiac catheter. Sensitivity of UCSDI with catheter was 4.7 μV/mA (R² = 0.999) in cylindrical gel (0.9% NaCl), and 3.2 μV/mA (R² = 0.92) in porcine heart tissue. The AE signal was detectable more than 25 mm away from the source in cylindrical gel (0.9% NaCl). Effect of transducer properties on UCSDI sensitivity is also investigated using simulation. The optimal ultrasound transducer parameters chosen for cardiac imaging are center frequency = 0.5 MHz and f/number = 1.4. Last but not least, this dissertation shows the result of implementing the optimized ultrasound parameters in live rabbit heart preparation, the comparison of different recording electrode configuration and multichannel UCSDI recording and reconstruction. The AE signal detected using the 0.5 MHz transducer was much stronger (2.99 μV/MPa) than the 1.0 MHz transducer (0.42 μV/MPa). The clinical lasso catheter placed on the epicardium exhibited excellent sensitivity without being too invasive. 3-dimensional cardiac activation maps of the live rabbit heart using only one pair of recording electrodes were also demonstrated for the first time. Cardiac conduction velocity for atrial (1.31 m/s) and apical (0.67 m/s) pacing were calculated based on the activation maps. The future outlook of this dissertation includes integrating UCSDI with 2-dimensional ultrasound transducer array for fast imaging, and developing a multi-modality catheter with 4-dimensional UCSDI, multi-electrode recording and echocardiography capacity.

Mémoire numérisé par la Division de la gestion de documents et des archives de l'Université de Montréal

Several studies suggest that action potential duration (APD) alternans play an important role in initiation of arrhythmias, while less is known about the alternans of early phases of action potential (AP) and phase relation between the two. Transmembrane potentials recorded from swine and canine ventricles were analyzed to determine the correlation and phase relation between alternans of early and late phases of an AP. In both species, for activation intervals ≤ 400 ms, action potential amplitude (APA) alternans occurred≥ 50% of times when APD alternans occurred and vice versa, both were mostly in phase. Also, alternans of APA and APD were mostly in phase with alternans of maximal rate of depolarization. The correlation between alternans in early and later parts of AP, however, was variable between species; APD10 and APD90 alternans were out of phase 81 % versus 34 % in canines and swines. These observations suggest that ionic mechanisms underlying alternans of depolarization and early repolarization phases may be distinct from those underlying later phases of repolarization. Simulations conducted to see the spatiotemporal effect of phase behavior between these alternans show that out of phase behavior suppresses oscillations in wavelength and minimizes the chances of spatial discordance.

This mixed-method study investigated the effects of self-directed physiological monitoring on therapists anxiety. Ten therapists participated in a10-week physiological monitoring training sessions while monitoring respiratory sinus arrhythmia (RSA) and heart rate variability (HRV). The participants completed the state-trait anxiety inventory questionnaire after having a first, sixth, or tenth therapy session with a client. This was designed to monitor their state anxiety while working with clients. A series of paired sampled t-tests was conducted to assess changes in HRV, RSA, trait anxiety, and state anxiety. One significant result was found: the RSA of the therapist increased significantly. Correlations existed between the HRV of the therapist increasing and the trait anxiety of the therapist decreasing through RSA training sessions although they were not significant at the .05 level.

Understanding the mechanisms behind unexplained abnormal heart rhythms is important for diagnosis and prevention of arrhythmias. Many studies have investigated the mechanisms at organ, tissue, cellular and molecular levels. There is considerable information available from tissue level experiments that investigate local action potential properties and from optical imaging to observe activity propagation properties at an organ level. By combining those electrophysiological properties together, in the present study we developed a simulation model that can help in estimation of the resulting body surface potentials from a specific electrical activity pattern within the myocardium. Some of the potential uses of our model include: 1) providing visualization of an entire electrophysiological event, i.e. surface potentials and associated source which would be optical imaging data, 2) estimation of QT intervals resulting from local action potential property changes, 3) aiding in improving defibrillation therapy by determining optimal timing and location of shocks.