Abstract
Due to the limited capability of the nerve cells to overcome injury, damage to the nervous system can cause paralysis, cerebral palsy, deafness, blindness, stroke; other functions can be lost through disease or amputation. If communication with the nervous system can be improved, information from sensors within the body could be used to control devices in aiding or ‘bridging’ damaged nerve areas and improving quality of life. One proposed method of performing this is through the development of sensors that can interact direct with the cells of the nervous system, and provide access to the signals produced by these nerve cells. This work presents a study of neural probes to quantify the effects of selected design factors in the ability of a probe to accurately sense information from a given nerve area. This was performed by examining a number of interwoven data sources, including neural recordings, histology of the recorded nerve, and analysis of the recordings characteristics (noise and impedance) of the recording electrodes, using a locust animal model, to assess different types of neural probes. The impedance of the electrodes was examined to develop a relationship between effective recording areas and impedances. Statistical tools allowed the study of the electrodes as single units as a function of their impedances, and to evaluate their performance. Mathematical developments helped to evaluate the ability of the sensors to detect the positions of the neurons with respect to the points of insertions, and therefore to assess their detection range and selectivity of recording. The results obtained from this work suggest that neural probes with electrode areas between 100 and 600 μm2 presented less susceptibility to noise and more to neural activity than larger or smaller areas. It also supports the argument that electrodes spacing higher than 50 pm increase the detection range.