Abstract
Abstract: :
Purpose: To develop and test a stimulus pattern generator as retina encoder (RE) module for animal experiments and RE tuning procedures of learning retina implants in humans including event tests and selectivity tuning. Methods: A PC-based stimulus pattern generator (SMG) was developed as software packet for generation of parallel electrical stimulation commands for microcontact arrays with up to several hundred electrodes in order to mimic various visual stimuli. A SMG input interface (II) was developed for initiation of stimulation events and monitoring of stimulation results. A SMG output interface (OI) was developed for management of output signals to a wireless signal transmitter and / or an array of constant current bi-phasic pulse generators (both provided by Fraunhofer Inst. IMS, Duisburg). A control interface (CI) as graphic user interface was designed to select various stimulation pattern parameters such as: electrode address, electrode-specific stimulation pulse profile (SP), time course of stimulation impulse rate (SR(t)), and spatio-temporal patterns for 'moving' stimulation events (MS(t)). Tests were performed for 3x3 and 5x5 arrays of microcontacts spatially spread over fixed grids as basis for the various spatial patterns. Results: (1) CI-selected, individual stimulation events for each of the electrodes were successfully elicited with stationary impulse rates SR(t) up to 1.000 imp/sec or with SR(t) modulation as brief bursts. (2) Moving MS(t) such as grids were successfully generated for an array of 5x5 microcontacts with individually selected SPs and SRs(t) with repetition rates (RR(t)) of at least 1.000 rep/sec at SR(t) of 250 imp/sec. (3) The shape of spatial patterns could be changed (morphed) with time. Conclusions: A. SMG is an important RE-module for tests in animals. B. SMG is required for event tests as baseline measurements of retina implants in humans. C. SMG is suitable for the selection of neighboring electrode clusters to provide selectivity tuning of RE in humans.
Keywords: pattern vision • computational modeling • retinal connections, networks, circuitry