0739-5175/05/$20.00©2005IEEE O ne of the frontiers in the biomedical sciences is the development of prostheses for the central nervous system (CNS) to replace higher thought processes that have been lost due to damage or disease. Prosthetic systems that interact with the CNS are currently being developed by several groups , though virtually all other CNS prostheses focus on sensory or motor system dysfunction and not on restoring cognitive loss resulting from damage to central brain regions. Systems designed to compensate for the loss of sensory input attempt to replace the transduction of physical energy from the environment into electrical stimulation of sensory nerve fibers (e.g., a cochlear implant or artificial retina) or the sensory cortex –. Systems designed to compensate for the loss of motor control do so through functional electrical stimulation (FES), in which preprogrammed stimulation protocols are used to activate muscular movement , , or by decoding premotor/motor cortical commands for the control of robotic systems –. The type of neural prosthesis that performs or assists a cognitive function is qualitatively different from the cochlear implant, artificial retina, or FES. We consider here a prosthetic device that functions in a biomimetic manner to replace information transmission between cortical brain regions , . In such a prosthesis, damaged CNS neurons would be replaced with a biomimetic system comprised of silicon neurons. The replacement silicon neurons would have functional properties specific to those of the damaged neurons and would both receive as inputs and send as outputs electrical activity to regions of the brain with which the damaged region previously communicated (Figure 1). Thus, the class of prosthesis being proposed is one that would replace the computational function of the damaged brain and restore the transmission of that computational result to other regions of the nervous system. Such a new generation of neural prostheses would have a profound impact on the quality of life throughout society; it would offer a biomedical remedy for the cognitive and memory loss accompanying Alzheimer’s disease, the speech and language deficits resulting from stroke, and the impaired ability to execute skilled movements following trauma to brain regions responsible for motor control. The Hippocampal System: Basis for Long-Term Declarative Memory We are in the process of developing such a cognitive prosthesis for the hippocampus, a region of the brain involved in the formation of new long-term memories. The hippocampus is responsible for what have been termed long-term declarative or recognition memories –: the formation of mnemonic labels that identify a unifying collection of features (e.g., those comprising a person’s face) and form relations between multiple collections of features (e.g., associating the visual features of a face with the auditory features of the name for that face). It is the degeneration and malformation of hippocampal neurons that is the underlying cause of the memory disorders associated with Alzheimer’s disease. Similarly, hippocampal pyramidal cells, particularly those in region CA1, are highly susceptible to even brief periods of anoxia, such as those that accompany stroke. Even blunt head trauma has been shown to be associated with a preferential loss of hippocampal neurons in the hilus of the dentate gyrus. Finally, there is a long history of association between hippocampal dysfunction (particularly in region CA3) and epileptiform activity. Thus, there is a wide array of neural damage and neurodegenerative disease conditions for which a hippocampal prosthesis would be clinically relevant. The hippocampus comprises several different subsystems that form a closed feedback loop (Figure 2); input from the neocortex enters via the entorhinal cortex, propagates through the intrinsic subregions of hippocampus, and then returns to the neocortex. The intrinsic pathways consist of a cascade of excitatory connections organized roughly transverse to the longitudinal axis of the hippocampus. As such, the hippocampus can be conceived of as a set of interconnected, parallel circuits , . The significance of this organizational feature is that, after removing the hippocampus from the brain, transverse slices (400 μm thick) of the structure may be maintained in vitro that preserve a substantial portion of the intrinsic circuitry.