R.
M. E. Sabbatini, PhD
State University of Campinas, Campinas/SP, Brazil
Artificial neural networks can represent and handle information by
means of interconnected processing elements, in a manner similar to
that of biologic neurons. Artificial neural networks are useful in
processing time-patterned biologic signals, such as electrocardiograms
and electroencephalograms, particularly in tasks involving pattern
classification and recognition. They do not need to be programmed,
since they are capable of learning. Using artificial neural networks to
build associative memories and perform true parallel processing permits
"intelligent" biomedical instrumentation.
Return to Home Page | Return to Abstracts Index |