Novel heuristics of functional neural networks: implications for future strategies in functional neurosurgery

A hypothesis is proposed that (a) the skeletomotor basal ganglia-thalamocortical loop functions as a model of the behavior of the body and the environment, and that (b) dopaminergic neurons of the substantia nigra pars compacta comprise the substrates of an error distribution system projecting to the striatum. This error signal initiates the learning process in the basal ganglia - learning starts with increasing intensity of the error signal and is complete when the signal is minimized. Parkinson's disease (PD) may be considered as a disruption of learning processes in the basal ganglia that results from progressive degeneration of the substrate that is the error distribution system for this functional motor loop. Numerous clinical and experimental observations obtained from functional procedures for PD that show identical clinical effects in alleviating parkinsonian symptoms, e.g. thermocoagulative lesions and chronic stimulation, can be explained through the use of this conceptual theory of basal ganglia function. Because any controlling neural network must possess a model of the behavior of its controlled object, the heuristics outlined in this theory are broadly applicable for explaining the function of the nervous system, as well as being useful for planning surgical procedures and future strategies in functional neurosurgery.