Human Body Communication Channel Characterization for Leadless Cardiac Pacemakers Mirko Maldari †,* , Karima Amara † , Ismael Rattalino † , Chadi Jabbour * and Patricia Desgreys * † Microport CRM, 4 Avenue R´ eaumur, 92140 Clamart, France * Institut Mines-T´ el´ ecom, T´ el´ ecom Paristech, LTCI, 46 Rue Barrault, 75013 Paris, France Email: mirko.maldari@crm.microport.com Abstract—Leadless Cardiac Pacemaker (LCP) devices are the cutting edge technology for cardiac rhythm management. An LCP requires to communicate with an external programmer and with different LCP nodes placed in different heart chambers. Human Body Communication (HBC) is an ultra-low power telemetry that exploits body conduction properties to propagate signals. There is a lack of studies in the literature about intra cardiac HBC channel characterization. Channel modeling is fundamental to prototype transceivers able to link different LCP devices. Quasi-Static Simulations based on Finite Element Method (FEM) have been used to characterize all HBC channels involved in LCP telemetry. A very accurate 3D model of a Human Torso has been designed and used to characterize HBC attenuation levels in a frequency range between 40 KHz and 20 MHz. This kind of approach will help to minimize animal trials for a more durable and optimized development. I. I NTRODUCTION In the last 60 years, electronics downscaling has conducted pacemaker devices to reach very low dimensions. Leadless Cardiac Pacemaker (LCP) is the last evolution in this field. It consists of a small capsule, anchored in one of the heart chambers, that can integrate all the functionalities of standard pacemakers. The technological bottleneck that prevents LCP to completely replace standard pacemakers stands in power budget limits for communication purposes [8]. Human Body Communication (HBC) is an ultra-low power communication that is a good candidate to overcome telemetry power bud- get issues. HBC is a conductive communication based on Ohmic propagation of currents through body tissues. Indeed an electrode pair is used to build up an electric field that propagates through Human Body reaching a second electrode pair used to receive the signal [14]. Moreover, HBC does not radiate meaning that it does not require an additional antenna, reducing the size of the device, and prevents eavesdropping since direct contact with the body is strictly required. Most of the academic studies about HBC channel modeling are focused on body-surface communications [1], [10], leading IEEE to include HBC as a Physical layer in IEEE 802.15.6 standard for Wireless Body Area Networks (WBAN). From the dawn of HBC communication, several studies have been performed in order to characterize the human body as a communica- tion channel. HBC channel modeling can be divided in two main categories: electric circuit modeling and computational electromagnetic modeling. Electric circuit models are simpli- fied solutions where human body is represented by a 4-port network made of bio-impedances. Computational Electromag- netic Modeling (CEM) is a computer aided methodology based on analytical or numerical analysis methods. A first approach of HBC channel modeling for implant devices was proposed in [11] where a boundary conditioned problem is analytically solved for HBC communication in a limb. The study shows the importance of the model geometry despite the fact that an approximated body model was used. The work proposed in this paper is based on Finite Element Method (FEM) simulations using a very accurate Torso model to characterize HBC channels involved in LCP applications. In particular, three main channels will be dealt in this paper: intra-cardiac channel between right chambers, Right Ventricle (RV) toward Left Ventricle (LV) and from RV to Body Surface (BS). II. NEW ACCURATE TORSO MODEL A. Quasi-Static Simulations The Finite Element Method is one of the most reliable numerical methods, it has been used for numerical com- putation of partial derivative equation systems for almost a century. COMSOL is a Multiphysics software for modeling and simulating complex physics systems. In order to reduce computational cost for low frequency problems COMSOL Multiphysics 5.3 employs a solver called AC/DC module that is based on Electro-Quasi-Static assumptions. Indeed it solves a current continuity equation problem using the FEM given voltage boundary conditions as described in equation system (1). It is worth recalling that quasi-static assumptions are a good approximation if the observation distance is at least ten times lower than the wavelength. LCP communication channels have distances that do not exceed 12 cm. Therefore, it was decided to limit the study at a maximum frequency of 20 MHz since the wavelength in the blood starts to be comparable with channel distances. The low frequency limit must be also considered in order to avoid interferences with electro-physiological signals and it was set at 40 KHz.      ∇· J = Q j,V J = σE + jωD + J e E = −∇V (1) B. Finite Element Model Development Human Body is a complex anisotropic medium. Therefore, a very accurate model has been used to characterize HBC