The overall goal of this instrument development for Systems Neurophysiology is to create functional devices to enable recording from very large ensembles of neuronal spike trains in behaving animals. The plan is to develop and integrate three tasks to be done in parallel. First, a new fabrication process will be employed to create thin film silicon recording probes, each with sixteen sites. The probe shapes and surface recording site properties will be optimized for long-term chronic recording to match the effectiveness of microwire arrays previously perfected by the P.I. Different spacing and sizes of surface recording sites will be patterned on the probe and different recording surfaces will be tested. A novel technology for fabrication of flexible microcables and multichip modules will be adapted to lead signals from the recording probes to external signal processing electronics. A second task is to create a design for a VLSI microchip to detect and amplify- signals from 16 channels. and multiplex signals to a single output lead. The chip will also provide switches with control to apply test/stimulation to a selected input. Later goals will be to produce a multisite probe with integrated active electronics and to create multiple probe assemblies. The processes to be developed are exceptionally favorable for commercial production methods. A low-cost, high-volume commercial product will be facilited through later participation of the staff at the Microelectronic Center of North Carolina (MCNC), Research Triangle, NC. A third task is to assemble a demultiplexer with control hardware/software and an advanced digital signal processing capability will be added to an existing workstation to support spike sorting from 256 channels or more. All three tasks will be done in parallel to incrementally add function to an existing, usable system. This complex, integrated project is made more feasib le by preliminary advances supported by NSF and elsewhere over past decade. This project aims to bring closur,9/to a decade-long development effort of the P.I. and his associates. The aim was to create and demonstrate the potential large scale neuron ensemble recording for decoding of information in patterns within distributed neuronal activity. Multivariate statistical methods have been applied to process the large volumes of data yielded by the new data acquisition methods. Recent publications by this group and others on multichannel recordings have created enthusiasm and momentum toward further advances in multichannel recording and analysis instrumentation. Advances in this capability are critical to enable a basic paradigm shift in the study of conceptual issues in the study of physiological neuronal networks.
