Computer aided identification of eugonic gram negative bacilli. Clustering for bacteriological screening

被引：1

作者：

Grenander, U. ^{[1
]}

Porres, J.M. ^{[1
]}

机构：

[1] Cent. Computer Informat. Sci., Brown Univ., Providence, R.I. 02912, United States

来源：

Computers in Biology and Medicine | 1976年 / 6卷 / 02期

基金：

美国国家科学基金会;

关键词：

PATTERN RECOGNITION SYSTEMS - STATISTICAL METHODS;

D O I：

10.1016/0010-4825(76)90041-X

中图分类号：

学科分类号：

摘要：

A system based on pattern recognition in terms of operationally meaningful criteria has been designed for the identification of eugonic gram-negative bacilli. It includes the selection of a battery of tests to be used as a screening. The results of these tests place each isolate in any one of a number of clusters. A second set of tests is used to identify the isolate as one of the species in the cluster. The system can identify most common species economically and can be used in clinical laboratories not equipped with computers. The approach described here uses statistics in order to determine a good choice of tests.

引用

页码：99 / 109