ABSTRACT
This chapter provides an overview of current technological trends, issues, and possibilities that have to be considered for successful integration of neuro-fuzzy algorithms. It covers the hardware specification requirements for the implementation of autonomous, safety-critical systems. The chapter describes the main hardware families implementing neural networks and fuzzy paradigms on-chip. The system architectures mostly reflect the conventional fuzzy-based strategy; that is to say, the fuzzifier interface driving the inference engine, which, in turn, drives the defuzzifier interface, to build up a structure targeted to efficiently process approximate reasoning and decision making algorithms. It includes the definition of the project specifications, to which all further decisions are related. The successful hardware integrated circuit implementation of safety-critical biomedical systems entails specific requirements, which increase the reliability of electronic systems and the patients comfort and autonomy, as well as reducing the overall costs of medical treatment.
