Mixed neural-conventional processing to differentiate airway diseases by means of functional non-invasive tests

Author

Parvis, M. ; Gulotta, C. ; Tochio, R.

Author_Institution

Dipt. di Elettronica, Politecnico di Torino, Italy

Volume

1

fYear

1999

fDate

1999

Firstpage

93

Abstract

This paper describes a processing technique that can be used to combine the pieces of information coming from different medical analyses. Such a technique is based on a mixed neural-and-conventional processing that allows both an easy neural network training and a robust estimation to be obtained. The paper is focused on the differentiation of asthma, bronchitis and emphysema by using functional non-invasive tests only, but the proposed technique can be easily applied to several different situations

Keywords

case-based reasoning; diseases; estimation theory; learning (artificial intelligence); lung; multilayer perceptrons; physiological models; pneumodynamics; uncertainty handling; airway diseases differentiation; asthma; binary training; bronchitis; bronchodilation; emphysema; expected uncertainty; functional noninvasive tests; guard neuron; lung diseases; mixed neural-conventional processing; multilayer perceptron; neural network training; pathology evidence index; respiratory parameters; robust estimation; spirometric data; winner takes all; Air pollution; Diseases; Information analysis; Lungs; Neural networks; Pathology; Performance evaluation; Robustness; Testing; Uncertainty;

fLanguage

English

Publisher

ieee

Conference_Titel

Instrumentation and Measurement Technology Conference, 1999. IMTC/99. Proceedings of the 16th IEEE

Conference_Location

Venice

ISSN

1091-5281

Print_ISBN

0-7803-5276-9

Type

conf

DOI

10.1109/IMTC.1999.776726

Filename

776726