Brain Theories, Applications 2
Memory Formation

The Interconnected Node (IN) brain model presented here does a fairly good job of handling the basic aspects of memory formation as described by Rolls. As he shows, many regions of the cerebral cortex, particularly the frontal lobes, have projections into the hippocampus. That area consists of many pyramid and mossy cells connected so as to tie together any pair (or group) of cortical regions which became active during a short time interval of each other. As long as these pairs (or groups) of circuits are kept connected, synapse growth can occur so as to interconnect the circuits in a more permanent manner.

A second possible aspect of this circuit interconnection, which I have not yet seen described elsewhere, concerns the role played by the entorhinal cortex. Rolls comments that the signals leading to the hippocampus from many other cortical regions pass through a relay station in the entorhinal cortex, located just behind the hippocampus, also on the inner edge of the cortical sheet, on the lower side of the temporal lobes. In the material cited here, he does not mention the function of these relay stations. However, Routtenberg notes two effects of electrical stimulation during the time memories are being formed. Disruptive stimulation of the hippocampus during the occurrence of the original event will prevent the formation of memories of that event. Disruptive stimulation of the entorhinal cortex at that time will not interfere with memory formation, but EC stimulation some time later will. Again, Routtenberg does not suggest a reason for this pattern.

I offer the following hypothesis. Either because of capacity limitations or timing questions, the hippocampus cannot hold the cortical circuits in contact long enough to allow synapse growth to occur. Instead, the entorhinal cortex serves as a secondary switch, providing a longer term hold on the circuit interconnections. There is a problem, however, with this interpretation. The cortical regions have connections both to and from the hippocampus, but only the incoming fibers are routed to the entorhinal cortex. Outgoing fibers make similar relay connections in the parahippocampal gyrus and in the perirhinal cortex. Rolls did not mention any cross connections between these latter areas and the entorhinal cortex. This would suggest an opposite interpretation; that the entorhinal cortex still has the task of providing a longer term hold, but that it does so in order to keep the circuits within the hippocampus active longer than the cortical circuits themselves stay active. Both of these interpretations are easily modelled by the Interconnected Node brain model.

A problem the IN brain model has in representing memory formation is its dependency on the chemical substrate. Dopamine has clearly been shown to play a major role in memory formation. Such effects could not easily be modelled with the current IN formulation.

Show list of book reviews sorted:
by Topic | by Title | by Author
Show list of essays sorted:
by Topic