Digital approach for cochlea's stimulation : a programmable micro stimulator driven by a flexible speech processing

We describe in this work a digital approach for cochlear stimulation. This would concern the design of an electronic micro-stimulator as well as the speech processing dedicated to drive this device. The design was versatile and numerical, that's why this under-the-skin micro-stimulator could be adapted to any external sound analyzer that could be driven by a digital processor 'DSP'. The design includes a transmission bus for differentiating two main stages: the decoding stage and the stimulation stage. The electronic circuit was then built around a logical processing unit that pilots the stimulation stage. After processing sounds by the external sound analyzer, appropriate numerical data would be transmitted to this internal micro-stimulator through a communication link mounted around an inductive coupling. The main functions assured during internal processing permitted to determine with great flexibility the stimulation current level to generate at each specified channel as well as the stimulation rhythm. The proposed speech processing consisted in filtering the acoustical signal by using one or the combine of these two algorithms: The FFT algorithm and the FIR-filter bank algorithm. This filtering was in fact a sounds' energy extraction that served for estimating stimuli shape. It permitted not only to process speech signal with great flexibility for delivering appropriate stimuli, but also to facilitate clinical adjustments. Its digital approach permitted to be adaptable to any apparatus driven by a 'DSP'.