In the case of spinal cord injurie, Functional Electrical Stimulation, which creates artificial muscle contractions by electrical stimulation, has proven to be capable of helping to restore impaired biological functions. Sensory information from natural sensors going through afferent nervous fibers could be used as feedback information in a closed-loop FES system. In order to get this information, we propose a new multipolar cuff electrode with hexagonal patterns as well as first elements of signal processing and amplification. The hexagonal cuff electrode presented in this paper allows higher sensitivity and a better rejection of parasitic signals than a classical tripolar cuff electrode. The preamplifier associated with this electrode presents the asset to integrate, at the transistor levels, a rejection technique of parasitic signals based on the average. This technique is used to perform electroneurogram (ENG) recording. This integration at the transistor level allows us to reduce the noise level due to processing electronics during sensory information acquisition.