RESEARCH

The mechanisms that lead to age- and sex-dependent changes in the heart pacemaker function
Age-associated heart pacemaker deterioration affects the health of many seemingly healthy elderly individuals and of patients with cardiac conditions. We showed that age-associated decrease in cAMP level is accompanying with a reduction in the spontaneous pacemaker beating rate and that an increase in the cAMP level can reverse this effect. With the ultimate goal of reversing age-associated pacemaker dysfunction, the proposed research program aims to understand the mechanisms that lead to age-associated deterioration of cAMP levels, to understand if cAMP deterioration is global or compartmentalized and to determine whether the mechanisms that lead to cAMP deterioration are sex-dependent.
In vitro platform to detect the effects of antiarrhythmic drugs on rate and rhythm in human cardiac tissue
Atrial fibrillation (AF) is the most common sustained arrhythmia and a significant risk factor for stroke. The mechanisms underlying AF episodes of different etiologies are not known. As a result, most antiarrhythmic drugs are ineffective. In addition, debate exists in the medical field whether treatment should target the heart rate (i.e., average heart beat interval) or rhythm (i.e., variability among heartbeat interval). We recently developed a method to measure heart rate and rhythm and link them with changes in intracellular processes in atrial and sinoatrial node tissue. Alterations in these activities were found in AF patients. Here, we aim to apply these methods in human atrial tissues and atria-derived human inducible pluripotent stem cells (hiPSC) to test antiarrhythmic drugs. We will focus on drugs that target Ca2+ and PKA signaling. In addition, new artificial intelligence (AI) methods based on time-series modelling will be applied to determine the toxicity dose of tested drugs, to predict the effect of drug combinations on heart rate and rhythm and to predict the effect of untested drugs based on responses of other drugs indicated for the same AF etiology. This platform will serve as a fast and affordable means of screening antiarrhythmic drugs with known targeted mechanisms, and will ultimately reduce morbidity and mortality.

Smartphone app-based AI algorithm for automatic analysis of 12-lead ECG

Cardiovascular diseases are currently the leading cause of global mortality.1 The official method to diagnose heart disease is to use the 12-lead ECG. Currently, only experienced cardiologists can perform the analysis while the measurements can be performed in any clinic. There are tasks even cardiologists cannot do, such as identifying intermittent arrythmia given normal sinus rhythm segments. We developed in the last two years an AI-based algorithm to diagnose the 14 most prevalent heart diseases and started its integration into a smartphone application. The algorithm is scalable disease-wise and specific population-wise, because, given sufficient training data, the approach assigns every target domain with a dedicated deep neural network-based binary classifier model. However, despite its accuracy, scalability and generality, major challenges (each, by itself, an open research question) are preventing its routine use in the clinic. The challenges we propose to work on include explainability of the system decisions, handling unknown 12-lead ECG formats, analysis of cardiac diseases from populations the system did not learn on (e.g., pediatric, or advanced age), identification of artifacts in the 12-lead ECG data (e.g., incorrect lead placement), identification of rhythmic diseases in normal sinus rhythm segments (NSRs), and lack of sufficient annotated training data for most diseases. We aim to use modern AI tools to overcome these challenges and prospectively test the new app in the hospital, validating its performance by cardiologists. Once validated, the community of cardiologists that use the app will collect a much larger and annotated dataset than was ever available, which, in turn, will be used to make our diagnosis more accurate and expand it to all heart diseases. This will enable early diagnosis of heart conditions via automated screening of the general population, ultimately reducing morbidity, mortality, and associated costs.

SYStematic approach to LINK RYR2 dysfunction WITH ArrhythmogenIC TRIGGERS in ATRIAL AND SINOATRIAL NODE

This project is a collaboration between the laboratories of Dr. Yael Yaniv, Associate Professor in Bioengineering at Technion University, Haifa, Israel, and Héctor Valdivia, Professor of Cardiovascular Medicine at the University of Wisconsin, Madison, USA.  The Valdivia lab has generated transgenic mice and rabbits harboring mutations associated with human catecholaminergic polymorphic ventricular tachycardia (CPVT) that faithfully recapitulate the main clinical features of the disease. Although some studies have shed light on the mechanisms by which CPVT mutations trigger ventricular arrhythmias, little is known on the mechanisms by which CPVT mutations affect the sinoatrial node function, the heart primary pacemaker. Because Yaniv’s lab expertise on sinoatrial node function in health, disease and aging, we are joining forces to investigate how two different CPVT mutations (one representing a gain-of-function and the other a loss-of-function of the ryanodine receptor, the main culprit of CPVT) affect the pacemaker of the heart and lead to fatal arrhythmias. We propose an in-depth study to determine the molecular mechanisms that link these CPVT mutations and the development of tachyarrhythmia in response to stress. This knowledge is clearly useful in rationalizing therapies to prevent and eliminate arrhythmogenic episodes in CPVT and other Ca2+-triggered arrhythmias.

Ventricular fibrillation event prediction

Ventricular fibrillation (VF) is a lethal cardiac arrhythmia and the cause of cardiac arrest in the vast majority of sudden cardiac death cases. Survival rates among victims can reach up to 90% in patients who receive immediate treatment; however, these rates decrease linearly by 10% for every minute of delay in treatment. So far, treatments are provided to patients only after the VF event; this is usually too late, when brain damage has already occurred. In addition, while sudden cardiac death can be treated with implantable defibrillators (ICDs), the majority (80%) of lethal arrhythmias occur in relatively low-risk patients, for whom the risk-benefit ratio of ICD implantation is not favorable. Early prediction of VF events could shorten the treatment delay, improve survival rates and clinical outcomes, and serve as a window for preventive therapy. To date, there is no effective clinical method for early prediction of VF events.

In my laboratory, we have developed a technology based on observing short (a few beats) and long (hundreds of beats) temporal patterns of heart rate changes, which, combined with machine learning tools, enables prediction of an event occurring within the next 24 hours. The method is based on heart rate as measured from ECG signals but can also be implemented using any device that provides heart rate signals (smart watches, fitness bands, or oxygen saturation monitors). Acquiring signals from such devices is inexpensive and non-invasive. The method was tested retrospectively on 84 patients with various arrhythmias and healthy volunteers. A patent on this technology has been filed by the Technion.

In this study, we will address the remaining gaps necessary to secure transition the product into industry. We will test the ability to detect a lethal arrhythmia in real time and assess the feasibility of doing so using a pulse wristband. We will evaluate the method in individuals who have an implanted defibrillator and can therefore be monitored for the occurrence of an event (they experience about 2–3 events per year), as well as in cardiac intensive care patients who are at high risk (on average, there are two events per week in the unit). By the end of the project, we expect to demonstrate that the lethal arrhythmia prediction technology works in real time. Based on this method, it will be possible to develop a smart watch or wristband that opens a new niche in this growing market—devices designed to provide an alert for a lethal arrhythmia event, not only detect its occurrence. A population of individuals at high risk for heart disease will be the first potential target market for this product.