THE CONCEPT OF DECISION POINTS

Conclusions

1. THE CONCEPT OF DECISION POINTS, COMBINED WITH THE METHOD OF MOMENTS, IS A USEFUL TOOL FOR ANALYZING THE COMPUTATIONAL CAPABILITIES OF REALISTIC DENDRITIC MODELS, IN WHICH SOME KINDS OF NONLINEARITIES ARE INCLUDED.

2. USING THE METHOD OF MOMENTS AND THE CONCEPT OF DECISION POINTS, WE SHOW THAT FROM THE SOMA VIEW POINT, THEORETICAL PASSIVE DENDRITES, WHERE ONLY THE SOMA SERVES AS A "DECISION POINT", DO NOT ADD SIGNIFICANTLY TO THE COMPUTATIONAL CAPABILITY OF THE NEURON (COMPARED TO A "POINT NEURON"). THIS WAS SHOWN BY ANALYZING THE "COST" OF PLACING A SYNAPSE AWAY FROM THE SOMA.

3. WE SHOW THAT THE STRUCTURE OF THE DENDRITIC TREE PROVIDES FOR MANY DECISION POINTS WITH VARIOUS COMPUTATIONAL PROPERTIES. MOREOVER, THERE ARE DIFFERENT SYNCHRONIZATION REQUIREMENTS FOR INPUTS IN LOCAL VERSUS GLOBAL DECISION POINTS. WE SUGGEST THAT THE ROLE OF THE SPATIAL DISTRIBUTION OF SYNAPSES ON THE DENDRITIC TREE IS FOR MAKING "LOCAL COMPUTATION CLUSTERS".

4. THE THEORETICAL FRAMEWORK DEMONSTRATED HERE MAY BE USED FOR A DETAILED ANALYSIS OF RECONSTRUCTED DENDRITIC TREES AND FOR ANALYZING THE EFFECT OF DIFFERENT BIOPHYSICAL PROPERTIES ON THEIR INTEGRATIVE PROPERTIES.

Back to Contents

Hagai Agmon-Snir Home Page