Expert system holds amazing promise. Dr Jack Myers Internist 1 database was made up of 100,000 numbers and facts about the patient for more than 500 diseases. Internist 1 used the numbers to generate a score, the score was used to make diagnoses or ask the doctor for more information. Internist 1 was a as good as many physicians but not as good as Myers. Other criticized Internist 1 as a Jack-in-The-box mimic of Myers. Myers countered the comment saying “computer scientist can not write programs to organize and use information as well as humans”. At the same time MYCIN was being developed by a group from Stanford University led by Edward Feigenbaum. The group began developing a computer technique to diagnose infectious disease. The research team took 10 meningitis patients and distributed the patient information too MYCIN, a Stanford facility member specializing in infectious disease treatment, a resident, and a medical student. The research team compared diagnoses and actual treatment recommendations and concluded that Mycin recommendations and diagnosis were at least as good as the three humans. Mycin type application could have lower cost introduction too emergency waiting rooms and improve accuracy and recommendation speed for a doctor, who must hand increasing numbers of patients on demand. Suppose, in the future, a person has an injury, drives to the hospital, and walks into the waiting area. The individual is greeted by one or many verbally interactive robot reception machine like Kismet and this cog begins extracting and storing information about the patient include all the billing information using voice recognition and synthesis. After the basic billing information is gather the Mycin algoritm starts collecting information about individual’s injury or disease and uses this information to make a preliminary diagnosis and make decisions to escalate attention. Once the doctor meets with the patient, the cog reviews the medical chart with the doctor and provides recommendations along the way assisting the doctor in decision making. As new information is added too the medical chart, the cog may ask for additional clarification questions and change recommendations. The patient information recommendations by the cog can be reviewed by a central computer with collective knowledge. The central computer can be polled too determine consensus on diagnosis and recommendation. The doctor concurrence statistics could be weighted against the cog advice. The more times the cog is concurred with the doctor, the higher the cogs rating. If a doctor habitually ignores the recommendations and disagrees, this information is passed too the central system for evaluation and assessment of cogs ability. Poor performing cogs are discontinued or replaced. As the cog abilities improves, the doctor will be more inclined to trust the cogs recommendations and malpractice suites could receive the reasoning of both the cog and the doctor in determine possible negligence. Eventually, the cog will become a valued assistant to reduce time, improve quality of medical treatment, and provide universal medical recommendations, at lower costs too the people.[Learn More ...]

&Links