Neural-network classification of normal and Alzheimer's disease subjects using high-resolution and low-resolution PET cameras

J. S. Kippenhan, W. W. Barker, J. Nagel, C. Grady, R. Duara

Research output: Contribution to journalArticle

38 Citations (Scopus)

Abstract

Neural-network classification methods were applied to studies of FDG-PET images of the brain acquired from a total of 77 'probable' Alzheimer's disease and 124 normal subjects at two different centers. Methods: Classification performances, as determined by relative-operating- characteristic (ROC) analyses of cross-validation experiments, were measured for FDG PET images obtained with either a 15-mm FWHM PETT V or a 6-mm FWHM Scanditronix PC-1024-7B camera for various methods of data representation. Neural networks were trained to distinguish between normal and abnormal subjects on the basis of regional metabolic patterns. For both databases, classification performance could be improved by increasing the 'resolution' of the representation (decreasing the region size) and by normalizing the regional metabolic values to the value of a reference region (occipital region). Results: The optimal classification performance for Scanditronix data (ROC area = 0.95) was higher than that for PETT V data (ROC area = 0.87). Under Bayesian theory, the classification performance with Scanditronix data corresponded to an ability to change a pre-test probability of disease of 50% to a post-test probability of either 90% for a positive classification or 10% for a negative classification. Conclusion: This classification can be used to either strongly confirm or rule out the presence of abnormalities.

Original languageEnglish
Pages (from-to)7-15
Number of pages9
JournalJournal of Nuclear Medicine
Volume35
Issue number1
StatePublished - Jan 1 1994
Externally publishedYes

Fingerprint

Alzheimer Disease
Occipital Lobe
Aptitude
Reference Values
Databases
Brain

Keywords

  • Alzheimer's disease
  • fluorodeoxyglucose
  • neural network

ASJC Scopus subject areas

  • Radiological and Ultrasound Technology

Cite this

Neural-network classification of normal and Alzheimer's disease subjects using high-resolution and low-resolution PET cameras. / Kippenhan, J. S.; Barker, W. W.; Nagel, J.; Grady, C.; Duara, R.

In: Journal of Nuclear Medicine, Vol. 35, No. 1, 01.01.1994, p. 7-15.

Research output: Contribution to journalArticle

Kippenhan, J. S. ; Barker, W. W. ; Nagel, J. ; Grady, C. ; Duara, R. / Neural-network classification of normal and Alzheimer's disease subjects using high-resolution and low-resolution PET cameras. In: Journal of Nuclear Medicine. 1994 ; Vol. 35, No. 1. pp. 7-15.
@article{a564232793ee402b9315b59beb6c901c,
title = "Neural-network classification of normal and Alzheimer's disease subjects using high-resolution and low-resolution PET cameras",
abstract = "Neural-network classification methods were applied to studies of FDG-PET images of the brain acquired from a total of 77 'probable' Alzheimer's disease and 124 normal subjects at two different centers. Methods: Classification performances, as determined by relative-operating- characteristic (ROC) analyses of cross-validation experiments, were measured for FDG PET images obtained with either a 15-mm FWHM PETT V or a 6-mm FWHM Scanditronix PC-1024-7B camera for various methods of data representation. Neural networks were trained to distinguish between normal and abnormal subjects on the basis of regional metabolic patterns. For both databases, classification performance could be improved by increasing the 'resolution' of the representation (decreasing the region size) and by normalizing the regional metabolic values to the value of a reference region (occipital region). Results: The optimal classification performance for Scanditronix data (ROC area = 0.95) was higher than that for PETT V data (ROC area = 0.87). Under Bayesian theory, the classification performance with Scanditronix data corresponded to an ability to change a pre-test probability of disease of 50{\%} to a post-test probability of either 90{\%} for a positive classification or 10{\%} for a negative classification. Conclusion: This classification can be used to either strongly confirm or rule out the presence of abnormalities.",
keywords = "Alzheimer's disease, fluorodeoxyglucose, neural network",
author = "Kippenhan, {J. S.} and Barker, {W. W.} and J. Nagel and C. Grady and R. Duara",
year = "1994",
month = "1",
day = "1",
language = "English",
volume = "35",
pages = "7--15",
journal = "Journal of Nuclear Medicine",
issn = "0161-5505",
publisher = "Society of Nuclear Medicine Inc.",
number = "1",

}

TY - JOUR

T1 - Neural-network classification of normal and Alzheimer's disease subjects using high-resolution and low-resolution PET cameras

AU - Kippenhan, J. S.

AU - Barker, W. W.

AU - Nagel, J.

AU - Grady, C.

AU - Duara, R.

PY - 1994/1/1

Y1 - 1994/1/1

N2 - Neural-network classification methods were applied to studies of FDG-PET images of the brain acquired from a total of 77 'probable' Alzheimer's disease and 124 normal subjects at two different centers. Methods: Classification performances, as determined by relative-operating- characteristic (ROC) analyses of cross-validation experiments, were measured for FDG PET images obtained with either a 15-mm FWHM PETT V or a 6-mm FWHM Scanditronix PC-1024-7B camera for various methods of data representation. Neural networks were trained to distinguish between normal and abnormal subjects on the basis of regional metabolic patterns. For both databases, classification performance could be improved by increasing the 'resolution' of the representation (decreasing the region size) and by normalizing the regional metabolic values to the value of a reference region (occipital region). Results: The optimal classification performance for Scanditronix data (ROC area = 0.95) was higher than that for PETT V data (ROC area = 0.87). Under Bayesian theory, the classification performance with Scanditronix data corresponded to an ability to change a pre-test probability of disease of 50% to a post-test probability of either 90% for a positive classification or 10% for a negative classification. Conclusion: This classification can be used to either strongly confirm or rule out the presence of abnormalities.

AB - Neural-network classification methods were applied to studies of FDG-PET images of the brain acquired from a total of 77 'probable' Alzheimer's disease and 124 normal subjects at two different centers. Methods: Classification performances, as determined by relative-operating- characteristic (ROC) analyses of cross-validation experiments, were measured for FDG PET images obtained with either a 15-mm FWHM PETT V or a 6-mm FWHM Scanditronix PC-1024-7B camera for various methods of data representation. Neural networks were trained to distinguish between normal and abnormal subjects on the basis of regional metabolic patterns. For both databases, classification performance could be improved by increasing the 'resolution' of the representation (decreasing the region size) and by normalizing the regional metabolic values to the value of a reference region (occipital region). Results: The optimal classification performance for Scanditronix data (ROC area = 0.95) was higher than that for PETT V data (ROC area = 0.87). Under Bayesian theory, the classification performance with Scanditronix data corresponded to an ability to change a pre-test probability of disease of 50% to a post-test probability of either 90% for a positive classification or 10% for a negative classification. Conclusion: This classification can be used to either strongly confirm or rule out the presence of abnormalities.

KW - Alzheimer's disease

KW - fluorodeoxyglucose

KW - neural network

UR - http://www.scopus.com/inward/record.url?scp=0028127877&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0028127877&partnerID=8YFLogxK

M3 - Article

C2 - 8271062

AN - SCOPUS:0028127877

VL - 35

SP - 7

EP - 15

JO - Journal of Nuclear Medicine

JF - Journal of Nuclear Medicine

SN - 0161-5505

IS - 1

ER -