DESIGN OF A FLEXIBLE SYNTACTIC PATTERN RECOGNITION SYSTEM.

A syntactic pattern recognition system for recognition of patterns such as brain stem auditory evoked potentials (BSAEP) is described. Data is digitally low-pass filtered and reduced to a string of terminal symbols. A general purpose software shell converts a set of production rules for a regular grammar into a recognizer for that grammar in the form of a deterministic finite-state automaton. The automaton parses the symbol string and activates procedural code referred to as side-effect functions to extract additional information. Efficacy of a finite-state grammar which simplifies the digital filtering and performance of the general purpose shell are demonstrated using a normal BSAEP.