Abstract This paper systematically demonstrates how a new sequential pattern-recognition technique can be used for computer diagnosis assessment. According to this technique it is possible to identify the patient's disease on the basis of information obtained by sequential analysis of symptoms by which the disease is manifested. Since the symptoms are analysed sequentially, the information that is needed for obtaining a unique diagnosis can be effectively minimised. The techique also enables one to develop a special program specifying the optimal direction of clinical investigation. This program determines at any particular stage of the diagnostic procedure which kind of clinical tests would yield at any given moment the diagnostically most discriminative data and should be consequently carried out with the highest priority. The amount of time and the number of clinical examintations required for the identification of any particular disease can thus be effectively reduced.
[1]
G. Lodwick,et al.
Computer Diagnosis of Primary Bone Tumors
,
1963
.
[2]
J. Overall,et al.
Conditional probability program for diagnosis of thyroid function.
,
1963,
JAMA.
[3]
A. W. Templeton,et al.
The computer analysis and diagnosis of gastric ulcers.
,
1965,
Radiology.
[4]
H. Warner,et al.
A mathematical approach to medical diagnosis. Application to congenital heart disease.
,
1961,
JAMA.
[5]
P. Armitage.
Sequential Medical Trials
,
1961,
Biomedicine / [publiee pour l'A.A.I.C.I.G.].
[6]
K SILINK.
A NEW MATHEMATICAL APPROACH TO ENDOCRINOLOGICAL DIAGNOSIS.
,
1964,
Progress in biocybernetics.
[7]
Abraham Wald,et al.
Statistical Decision Functions
,
1951
.
[8]
Chi Hau Chen,et al.
A Note on Sequential Decision Approach to Pattern Recognition and Machine Learning
,
1966,
Inf. Control..