MYCIN was the 1970s expert system designed to advise physicians about bacterial infections and antibiotic treatment.
Developed at Stanford University beginning in the early 1970s, MYCIN used a collection of approximately 600 production rules. It asked clinicians questions about a patient, considered symptoms and laboratory findings, and generated recommendations about likely blood infections and suitable antibiotics.
MYCIN was notable for representing uncertainty with certainty factors rather than requiring every medical fact to be absolutely true or false. It could also explain aspects of its reasoning by identifying rules that supported its conclusions.
The system was an important demonstration of rule-based medical decision support, but it was not deployed routinely in hospitals. Concerns included legal responsibility, integration into clinical practice, and the need for reliable patient data. MYCIN’s techniques influenced later expert systems and knowledge-based AI research.