Information Technology Reference
In-Depth Information
system for identifying chemical structures from mass spectrograms; and MYCIN,
an experimental expert system to assist in the selection of antibiotic therapy for
patients with infectious diseases of blood and nervous system. A host of experi-
mental medical expert systems such as CASNET, INTERNIST, PIP, and others fol-
lowed. A new field of research and practice developed that has come to be known as
Artificial Intelligence in Medicine (AIM), Medical Artificial Intelligence, medical
knowledge-based systems research, or medical expert systems research. The aim
was to build computer programs that could provide the physician with diagnoses
and treatment suggestions. To achieve this goal, the initial method of knowledge
engineering used has been supplemented with several new techniques such as neu-
rocomputing, evolutionary programming, case-based reasoning, and fuzzy logic.
But despite tremendous progress made in the meantime, the products of AIM re-
search are not yet good enough to compete with expert physicians. So, the initial,
ambitious term 'medical expert system' is being more and more displaced by the
humble term 'medical decision support system'.
In my view, clinical AIM research is an experimental science of clinical practice
that in the long run will produce decision support systems that are much more com-
petent than the individual physician in making clinical judgments. It will utilize all,
or many, of the available systems of logic, mathematics, and computer sciences to
bring about the 'physician machine' (PM) that will gain supremacy over the intellec-
tual capacity and expertise of physicians. Physicians will serve as mobile assistants
of the PM by the end of this century at the very latest.” [72]
3.5
The Geometry of Fuzzy Sets as Points in a Hypercube
As we mentioned already in chapter 3.4.1, the concept of a “fuzzy algorithm”
opened the door to the first fuzzy application systems. However, when Zadeh pre-
sented this idea at the first time in 1968, he was aware of its coriousity. Usually
the success of algorithms depends upon precision. An algorithm must be com-
pletely unambiguous and error-free in order to result in a solution. The path to a
solution amounts to a series of commands which must be executed in succession.
Algorithms formulated mathematically or in a programming language are based on
classical set theory. Each constant and variable is precisely defined; every func-
tion and procedure has a definition set and a value set. Each command builds upon
them. Successfully running a series of commands requires that each result (output)
of the execution of a command lies in the definition range of the following com-
mand, that it is, in other words, an element of the input set for the series. Not even
the smallest inaccuracies may occur when defining these coordinated definition and
value ranges. But now, Zadeh saw “that in real life situations people think certain
things. They thought like algorithms but not precisely defined algorithms.” [106]
Inspired by this idea, he wrote: “Essentially, its purpose is to introduce a basic con-
cept which, though fuzzy rather than precise in nature, may eventually prove to be
of use in a wide variety of problems relating to information processing, control, pat-
tern recognition, system identification, artificial intelligence and, more generally,
