Design of a laboratory for multineuron studies

The authors describe a number of procedures necessary for the conduct of a multineuron experiment. They have identified three problem areas that are expected to be particularly noticeable as larger assemblies are studied for long periods under conditions where the subject brain is engaged in a behavioral or learning task. The first concerns monitoring a moving animal. It is necessary to develop better multielectrode systems as well as appropriate manipulators for the high degree of stability that is required. Multichannel amplification (perhaps multiplexing with telemetry) must be developed in sizes that can be carried on the head. Second, meaningful feedback must be given to the experiments in time to help modify the conditions of the experiment. Full cross correlation analysis of simultaneously recorded neurons takes a large amount of capacity and time, which slows experimentation. New reduction or abstraction measurement that is practical for rapid feedback is needed. Finally, a better description of a neuronal assembly is needed. A series of new tools deployed to carry out multineuron experiments are described, including electrodes and computer analysis techniques.