Recording Of Cortex Responses Evoked By Electrical Stimulation On Optic Nerve And Research Of Signal Processing Methods

Retinal diseases, such as retinitis pigmentosa (RP) or age-related macular degeneration (AMD) are leading causes of untreatable blindness. Micro-electronic based visual prostheses hold great promise for the treatment of such diseases. Since a large percentage of the inner retinal neurons remain intact in these patients, many research efforts have been made to develop retinal stimulation based on visual prostheses. The other two visual prostheses are optic nerve visual prosthesis and cortex visual prosthesis. In this thesis, a new visual prosthesis based on intraorbital optic nerve stimulation with penetrating electrodes was introduced. The main contents of this thesis included the following three aspects:(1) Introduced a new visual prosthesis based on intraorbital optic nerve stimulation with penetrating electrodes. Teflon-coated tungsten wire electrodes were inserted intraorbitally into the optic nerve. A charge-balanced biphasic current was delivered to the optic nerve via inserted wire electrodes in the anaesthetized rabbits. Stimulus pulse amplitude, duration and waveform were varied respectively, Electrically Evoked Potentials (EEPs) were recorded by epidural electrodes array placed over the visual cortex.(2) Researched in EEPs'time-frequency analysis. This thesis researched the EEPs recorded using different stimulus parameters with the methods of Short Time Fourier Transform (STFT), Wigner-Ville Distribution (WVD) and Wavelet Transform (WT), and compared each method's advantages and disadvantages.(3) Presented a method of eliminating artifacts of multi-channel EEPs evoked by electrical stimulation with Independent Component Analysis (ICA). Because the recording data can be regarded as the linear mixtures of the statistically independent artifacts and the EEPs, First, the multi-channel recording data were whitened and the dimension was reduced to realize the de-noising purpose. Then the EEPs and the artifacts were separated by using ICA algorithm. By comparing with the artifacts recorded after the optic nerve was blocked by lidocaine, it is proved that this method is effective.