New technical approaches to monitoring and interpreting the dynamics of real neural networks

Understanding the functional organization of large-scale real neural networks will require the development of techniques capable of predicting, recording, and interpreting the activities of the very large numbers of neurons that participate in their operation. This presents a considerable technical problem for systems-level neurobiologists and associated biomedical engineers. The authors describe their recent work using silicon manufacturing techniques to make multineuron in vivo recording devices, as well as the software tools developed for realistically simulating the neural networks studied.<>