Neural networks in radiology: An introduction and evaluation in a signal detection task

John M Boone, V. G. Sigillito, G. S. Shaber

Research output: Contribution to journalArticle

56 Citations (Scopus)

Abstract

Neural networks are a computer architecture, implementable in software or hardware, that allow an entirely new approach to the computerized perception of data. These so-called connectionist models are inspired by what is known about the architecture of biological neurons, in which the 'intelligence' or processing capability of the network is a result of the interconnection strengths between large arrays of nonlinear processing nodes. Neural networks are described and then are used to analyze the common radiological problem of pattern recognition on a noisy background. Classical signal detection theory is used to compare network performance against that of human observers, using computer-generated sets of very simple 'nodules'. The neural network performed with better accuracy, relative to human observer performance, in the detection of this elementary test object. Although these results may not scale up with more complex images, the favorable performance of neural networks at this level suggests that further investigation is warranted.

Original languageEnglish (US)
Pages (from-to)234-241
Number of pages8
JournalMedical Physics
Volume17
Issue number2
DOIs
StatePublished - 1990
Externally publishedYes

Fingerprint

Neural Networks (Computer)
Radiology
Computer Systems
Intelligence
Software
Neurons
Psychological Signal Detection
Recognition (Psychology)

Keywords

  • connectionist models
  • neural networks
  • receiver operating curve

ASJC Scopus subject areas

  • Biophysics

Cite this

Neural networks in radiology : An introduction and evaluation in a signal detection task. / Boone, John M; Sigillito, V. G.; Shaber, G. S.

In: Medical Physics, Vol. 17, No. 2, 1990, p. 234-241.

Research output: Contribution to journalArticle

Boone, John M ; Sigillito, V. G. ; Shaber, G. S. / Neural networks in radiology : An introduction and evaluation in a signal detection task. In: Medical Physics. 1990 ; Vol. 17, No. 2. pp. 234-241.
@article{269a283cfa6d439390f1b7fc38283333,
title = "Neural networks in radiology: An introduction and evaluation in a signal detection task",
abstract = "Neural networks are a computer architecture, implementable in software or hardware, that allow an entirely new approach to the computerized perception of data. These so-called connectionist models are inspired by what is known about the architecture of biological neurons, in which the 'intelligence' or processing capability of the network is a result of the interconnection strengths between large arrays of nonlinear processing nodes. Neural networks are described and then are used to analyze the common radiological problem of pattern recognition on a noisy background. Classical signal detection theory is used to compare network performance against that of human observers, using computer-generated sets of very simple 'nodules'. The neural network performed with better accuracy, relative to human observer performance, in the detection of this elementary test object. Although these results may not scale up with more complex images, the favorable performance of neural networks at this level suggests that further investigation is warranted.",
keywords = "connectionist models, neural networks, receiver operating curve",
author = "Boone, {John M} and Sigillito, {V. G.} and Shaber, {G. S.}",
year = "1990",
doi = "10.1118/1.596501",
language = "English (US)",
volume = "17",
pages = "234--241",
journal = "Medical Physics",
issn = "0094-2405",
publisher = "AAPM - American Association of Physicists in Medicine",
number = "2",

}

TY - JOUR

T1 - Neural networks in radiology

T2 - An introduction and evaluation in a signal detection task

AU - Boone, John M

AU - Sigillito, V. G.

AU - Shaber, G. S.

PY - 1990

Y1 - 1990

N2 - Neural networks are a computer architecture, implementable in software or hardware, that allow an entirely new approach to the computerized perception of data. These so-called connectionist models are inspired by what is known about the architecture of biological neurons, in which the 'intelligence' or processing capability of the network is a result of the interconnection strengths between large arrays of nonlinear processing nodes. Neural networks are described and then are used to analyze the common radiological problem of pattern recognition on a noisy background. Classical signal detection theory is used to compare network performance against that of human observers, using computer-generated sets of very simple 'nodules'. The neural network performed with better accuracy, relative to human observer performance, in the detection of this elementary test object. Although these results may not scale up with more complex images, the favorable performance of neural networks at this level suggests that further investigation is warranted.

AB - Neural networks are a computer architecture, implementable in software or hardware, that allow an entirely new approach to the computerized perception of data. These so-called connectionist models are inspired by what is known about the architecture of biological neurons, in which the 'intelligence' or processing capability of the network is a result of the interconnection strengths between large arrays of nonlinear processing nodes. Neural networks are described and then are used to analyze the common radiological problem of pattern recognition on a noisy background. Classical signal detection theory is used to compare network performance against that of human observers, using computer-generated sets of very simple 'nodules'. The neural network performed with better accuracy, relative to human observer performance, in the detection of this elementary test object. Although these results may not scale up with more complex images, the favorable performance of neural networks at this level suggests that further investigation is warranted.

KW - connectionist models

KW - neural networks

KW - receiver operating curve

UR - http://www.scopus.com/inward/record.url?scp=0025355166&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0025355166&partnerID=8YFLogxK

U2 - 10.1118/1.596501

DO - 10.1118/1.596501

M3 - Article

C2 - 2333049

AN - SCOPUS:0025355166

VL - 17

SP - 234

EP - 241

JO - Medical Physics

JF - Medical Physics

SN - 0094-2405

IS - 2

ER -