Improved Nerve Cuff Electrode Recordings by Sub-threshold Anodic Currents

Computer simulatioins indicate that the propagation velocity of action potentials in a length o f a nerve axon can be decreased by subthreshold extracellular anodic currents. This phenomenon can be used to increase the amplitude of whole nerve recordings madie with a short cuff electrode with circumferencial metal bands, since larger propagation delays between the bands result in larger recorded signals. Computer simulations predicting the slowing effect of anodic currents and the experimental data verifying, the simulations are presented. The increase in the amplitude of nerve signals (fivefold), recorded experimentally from a short cuff, is demonstrated. INTRODUCTION The signal amplitudes increase with increasing cuff lengths when a cuff electrode with circumferencial metal bands (as the contact points) is used to record peripheral nerve activity [ 11. This effect can be attributed to the increase in propagation delay between the bands. With short cuff lengths, it is mostly the slow ( i.e. small) fibers that contribute to the recordings. However, the maximd dissectable length of the nerve places a limitation on the cuff length in most in vivo studies. Computer simulations indicate that anodic currents decrease the propagation velocity of action potentials by hyperpolarizing the membrane and increasing the time required by the membrane to reach ,the stimulation threshold. Thus, the propagation delay between the recordings obtained from the individual band electrodes that are placed along the cuff increases, thereby making the c u f f appear longer than its physical size. Thus, we propose that anodic currents can be used to increase the amplitude of whole nerve recordings made with a short cuff electrode. METHODIS The effect of applied anodic cmlents on the propagation of action potentials of a single myelinated nerve fiber (an axon of 10 pm caliber and of 10 cm length, i.e. 101 nodes) is demonstrated by computer simulations (Figure 1). The simulations are done using MiURON [2]. The active behavior of the cell membrane al. each node of Ranvier is simulated using an active mammalian nerve model [3]. The internodal sections are simulated by a single intracellular resistivity element (Ra). The exlracellular voltage profile along the axon due to a monopolar electrode placed in an infinite conductive medium is calculated (Figure 2). Discrete current sources proportional to sampled version of the second order spatial difference of this volltage profile are applied to the nodes intracellularly and the nodes are connected together at the extracellular site. This was shown to be equivalent to applying discrete voltage sources at the extracellular site as the sampled version of the extracellular voltage profile [4]. A stimulation pulse (10 psec, 20 nA) is applied at the left end of the axon. The temporal waveform of the transmembrane voltage is calculated at each node as the action potential propagates. The conduction velocities are calculated between consecutive nodes by taking the reciprocal of the time delay between the peaks of the voltage waveforms. The cumulative delay is defined as the total time period from the time of stimulation to the peak of activity at a given node.