Automated recognition of lateral from PA chest radiographs

Saving seconds in a PACS environment

John M Boone, Greg S. Hurlock, J Anthony Seibert, Richard L. Kennedy

Research output: Contribution to journalArticle

10 Citations (Scopus)

Abstract

Images acquired in a two-view digital chest examination are frequently not electronically distinguishable. As a result the lateral and posterioanterio (PA) images are often improperly positioned on a PACS work station. A series of 1998 chest radiographs (999 lateral, 999 PA or AP) were used to develop a neural network classifier. The images were down-sampled to 16 × 16 matrices, and a feed-forward neural network was trained and tested using the "leave-one-out" method. Using five nodes in the hidden layer, the neural network correctly identified 987 of the 999 test cases (98.8%) (average of six runs). The simple architecture and speed of this technique suggests that it would be a useful addition to PACS work station software. The accumulated time saved by correctly positioning the lateral and PA chest images on the work station monitors in accordance with each radiologist's hanging protocols was estimated to be about 1 week of radiologist time per year.

Original languageEnglish (US)
Pages (from-to)345-349
Number of pages5
JournalJournal of Digital Imaging
Volume16
Issue number4
DOIs
StatePublished - Dec 2003

Fingerprint

Picture archiving and communication systems
Thorax
Neural networks
Feedforward neural networks
Classifiers
Software
Radiologists

Keywords

  • Chest radiography
  • Neural networks
  • Pattern recognition
  • Picture archiving and communication system (PACS)

ASJC Scopus subject areas

  • Radiology Nuclear Medicine and imaging
  • Radiological and Ultrasound Technology
  • Computer Vision and Pattern Recognition

Cite this

Automated recognition of lateral from PA chest radiographs : Saving seconds in a PACS environment. / Boone, John M; Hurlock, Greg S.; Seibert, J Anthony; Kennedy, Richard L.

In: Journal of Digital Imaging, Vol. 16, No. 4, 12.2003, p. 345-349.

Research output: Contribution to journalArticle

@article{752f540fa45b4d61830dbeafa72ea628,
title = "Automated recognition of lateral from PA chest radiographs: Saving seconds in a PACS environment",
abstract = "Images acquired in a two-view digital chest examination are frequently not electronically distinguishable. As a result the lateral and posterioanterio (PA) images are often improperly positioned on a PACS work station. A series of 1998 chest radiographs (999 lateral, 999 PA or AP) were used to develop a neural network classifier. The images were down-sampled to 16 × 16 matrices, and a feed-forward neural network was trained and tested using the {"}leave-one-out{"} method. Using five nodes in the hidden layer, the neural network correctly identified 987 of the 999 test cases (98.8{\%}) (average of six runs). The simple architecture and speed of this technique suggests that it would be a useful addition to PACS work station software. The accumulated time saved by correctly positioning the lateral and PA chest images on the work station monitors in accordance with each radiologist's hanging protocols was estimated to be about 1 week of radiologist time per year.",
keywords = "Chest radiography, Neural networks, Pattern recognition, Picture archiving and communication system (PACS)",
author = "Boone, {John M} and Hurlock, {Greg S.} and Seibert, {J Anthony} and Kennedy, {Richard L.}",
year = "2003",
month = "12",
doi = "10.1007/s10278-003-1662-y",
language = "English (US)",
volume = "16",
pages = "345--349",
journal = "Journal of Digital Imaging",
issn = "0897-1889",
publisher = "Springer New York",
number = "4",

}

TY - JOUR

T1 - Automated recognition of lateral from PA chest radiographs

T2 - Saving seconds in a PACS environment

AU - Boone, John M

AU - Hurlock, Greg S.

AU - Seibert, J Anthony

AU - Kennedy, Richard L.

PY - 2003/12

Y1 - 2003/12

N2 - Images acquired in a two-view digital chest examination are frequently not electronically distinguishable. As a result the lateral and posterioanterio (PA) images are often improperly positioned on a PACS work station. A series of 1998 chest radiographs (999 lateral, 999 PA or AP) were used to develop a neural network classifier. The images were down-sampled to 16 × 16 matrices, and a feed-forward neural network was trained and tested using the "leave-one-out" method. Using five nodes in the hidden layer, the neural network correctly identified 987 of the 999 test cases (98.8%) (average of six runs). The simple architecture and speed of this technique suggests that it would be a useful addition to PACS work station software. The accumulated time saved by correctly positioning the lateral and PA chest images on the work station monitors in accordance with each radiologist's hanging protocols was estimated to be about 1 week of radiologist time per year.

AB - Images acquired in a two-view digital chest examination are frequently not electronically distinguishable. As a result the lateral and posterioanterio (PA) images are often improperly positioned on a PACS work station. A series of 1998 chest radiographs (999 lateral, 999 PA or AP) were used to develop a neural network classifier. The images were down-sampled to 16 × 16 matrices, and a feed-forward neural network was trained and tested using the "leave-one-out" method. Using five nodes in the hidden layer, the neural network correctly identified 987 of the 999 test cases (98.8%) (average of six runs). The simple architecture and speed of this technique suggests that it would be a useful addition to PACS work station software. The accumulated time saved by correctly positioning the lateral and PA chest images on the work station monitors in accordance with each radiologist's hanging protocols was estimated to be about 1 week of radiologist time per year.

KW - Chest radiography

KW - Neural networks

KW - Pattern recognition

KW - Picture archiving and communication system (PACS)

UR - http://www.scopus.com/inward/record.url?scp=2942706065&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=2942706065&partnerID=8YFLogxK

U2 - 10.1007/s10278-003-1662-y

DO - 10.1007/s10278-003-1662-y

M3 - Article

VL - 16

SP - 345

EP - 349

JO - Journal of Digital Imaging

JF - Journal of Digital Imaging

SN - 0897-1889

IS - 4

ER -