COMP3330 History Assignment

Artificial Intelligence Applications In Medicine

Expert Systems

History Of Expert Systems In Medicine

The development of expert systems (also known as knowledge-based systems) in the late 1960s and early 1970s was a significant breakthrough in AI. This technology was soon applied to various tasks in the field of medicine.

A medical expert system contains large amounts of medical knowledge, usually about a very specifically defined task, and is able to reason with data from individual patients to come up with reasoned conclusions. Although there are many variations, the knowledge within an expert system is typically represented in the form of a set of rules.

The first expert system designed for a medical application was MYCIN, developed by Dr. Edward Shortliffe in 1975 as part of his doctorate thesis. He was assisted by Ed Feigenbaum and Bruce Buchanan (who had worked on the earlier DENDRAL expert system).

MYCIN was tested by human experts in the area of bacterial infections (including physicians), and was shown to provide accurate results. This success inspired the development of other medical AI systems, many of which were rule-based expert systems. Among the first of these systems included:

Causal-Associated Network (CASNET):
Written by scientists at the Rutgers Research Resource on Computers in Biomedicine in 1976. CASNET used physiological models for the diagnosis and treatment of eye diseases.
 

INTERNIST:
Written by Harry Pople and Jack Myers in the late 1970s and early 1980s. This system was a diagnostic aid that combined a large database of disease/manifestation associations with techniques for problem formulation. It investigated a broad range of problems in internal medicine.

Many similar rule-based systems have been developed for medical applications over the years. (A code sample from a simple rule-based expert system for diagnosing nasal traumas can be found here.) However, due to the extreme complexity of maintaining rule sets with more than a few thousand rules, rule-based systems have generally been devoted to narrow application areas.

Expert Systems In Medicine Today

Expert systems are the commonest type of medical AI system in routine clinical use today. One of the most successful areas in which expert systems have been applied is in the clinical laboratory (e.g. pathology applications). Other types of clinical tasks to which expert systems can be applied include:

Generating alerts and reminders:
In real-time situations, an expert system attached to a monitor can warn of changes in a patient's condition. In less acute circumstances, it may scan laboratory test results or drug orders and send reminders or warnings through an e-mail system if an evaluation of data shows possibly critical developments.
 

Diagnostic assistance:
When a patient's case is complex, rare or the person making the diagnosis is simply inexperienced, an expert system can help come up with likely diagnoses based on patient data.
 

Therapy critiquing and planning:
Systems can either look for inconsistencies, errors and omissions in an existing treatment plan, or can be used to formulate a treatment based upon a patient's specific condition and accepted treatment guidelines.
 

Image recognition and interpretation:
Many medical images can now be automatically interpreted, from X-rays through to more complex images like angiograms, CT and MRI scans. This is of value in mass-screenings, for example, when the system can flag potentially abnormal images for detailed human attention.

Examples of expert systems in clinical use include:

Perfex, a rule-based expert system:
This system aids in the diagnosis of heart disease and is currently undergoing clinical evaluation. The system infers the extent and severity of coronary artery disease from myocardial perfusion imaging and produces a report that summarizes the condition of the three main arteries. Perfex was developed at Georgia Tech using Blaze Software's Nexpert, an object-oriented development environment for rule-based expert systems.
 

Agilent Acute Cardiac Ischemia Time-Insensitive Predictive Instrument:
This device, produced by Agilent Technologies, applies an expert system knowledge base to determine the probability of patients having certain types of heart attacks. A study done by the Agency for Healthcare Policy and Research concluded that more than 200,000 unnecessary hospitalisations and 100,000 unnecessary cardiac care unit admissions could be prevented each year if this device were used in U.S. emergency rooms.
 

Figure 1: The Agilent Acute Cardiac Ischemia Time-Insensitive Predictive Instrument,
from Agilent Technologies, increases the accuracy of diagnosing acute cardiac ischemia.