Automated identification of Fos expression

D. Young, J. Ma, S. Cherkerzian, M.P. Froimowitz, D.J. Ennulat, B.M. Cohen, M.L. Evans, N. Lange

Research output: Contribution to journalArticle

Abstract

The concentration of Fos, a protein encoded by the immediate-early gene c-fos, provides a measure of synaptic activity that may not parallel the electrical activity of neurons. Such a measure is important for the difficult problem of identifying dynamic properties of neuronal circuitries activated by a variety of stimuli and behaviours. We employ two-stage statistical pattern recognition to identify cellular nuclei that express Fos in two-dimensional sections of rat forebrain after administration of antipsychotic drugs. In stage one, we distinguish dark-stained candidate nuclei from image background by a thresholding algorithm and record size and shape measurements of these objects. In stage two, we compare performance of linear and quadratic discriminants, nearest-neighbour and artificial neural network classifiers that employ functions of these measurements to label candidate objects as either Fos nuclei, two touching Fos nuclei or irrelevant background material. New images of neighbouring brain tissue serve as test sets to assess generalizability of the best derived classification rule, as determined by lowest cross-validation misclassification rate. Three experts, two internal and one external, compare manual and automated results for accuracy assessment. Analyses of a subset of images on two separate occasions provide quantitative measures of inter- and intra-expert consistency. We conclude that our automated procedure yields results that compare favourably with those of the experts and thus has potential to remove much of the tedium, subjectivity and irreproducibility of current Fos identification methods in digital microscopy.
LanguageEnglish
Pages351-364
Number of pages14
JournalBiostatistics
Volume2
Issue number3
DOIs
Publication statusPublished - Sep 2001

Fingerprint

Nucleus
Immediate-Early Genes
Prosencephalon
Antipsychotic Agents
Microscopy
Misclassification Rate
Shape Measurement
Neurons
Classification Rules
Brain
Dynamic Properties
Test Set
Thresholding
Discriminant
Cross-validation
Pattern Recognition
Artificial Neural Network
Neuron
Lowest
Nearest Neighbor

Keywords

  • Amygdala
  • artificial neural networks
  • digital image analysis
  • immediate-early gene protein
  • immunohistochemistry
  • microscopy
  • digital microscopy

Cite this

Young, D., Ma, J., Cherkerzian, S., Froimowitz, M. P., Ennulat, D. J., Cohen, B. M., ... Lange, N. (2001). Automated identification of Fos expression. Biostatistics, 2(3), 351-364. https://doi.org/10.1093/biostatistics/2.3.351
Young, D. ; Ma, J. ; Cherkerzian, S. ; Froimowitz, M.P. ; Ennulat, D.J. ; Cohen, B.M. ; Evans, M.L. ; Lange, N. / Automated identification of Fos expression. In: Biostatistics. 2001 ; Vol. 2, No. 3. pp. 351-364.
@article{519c53b5c1ee4027806803b6607e1334,
title = "Automated identification of Fos expression",
abstract = "The concentration of Fos, a protein encoded by the immediate-early gene c-fos, provides a measure of synaptic activity that may not parallel the electrical activity of neurons. Such a measure is important for the difficult problem of identifying dynamic properties of neuronal circuitries activated by a variety of stimuli and behaviours. We employ two-stage statistical pattern recognition to identify cellular nuclei that express Fos in two-dimensional sections of rat forebrain after administration of antipsychotic drugs. In stage one, we distinguish dark-stained candidate nuclei from image background by a thresholding algorithm and record size and shape measurements of these objects. In stage two, we compare performance of linear and quadratic discriminants, nearest-neighbour and artificial neural network classifiers that employ functions of these measurements to label candidate objects as either Fos nuclei, two touching Fos nuclei or irrelevant background material. New images of neighbouring brain tissue serve as test sets to assess generalizability of the best derived classification rule, as determined by lowest cross-validation misclassification rate. Three experts, two internal and one external, compare manual and automated results for accuracy assessment. Analyses of a subset of images on two separate occasions provide quantitative measures of inter- and intra-expert consistency. We conclude that our automated procedure yields results that compare favourably with those of the experts and thus has potential to remove much of the tedium, subjectivity and irreproducibility of current Fos identification methods in digital microscopy.",
keywords = "Amygdala, artificial neural networks, digital image analysis, immediate-early gene protein, immunohistochemistry, microscopy, digital microscopy",
author = "D. Young and J. Ma and S. Cherkerzian and M.P. Froimowitz and D.J. Ennulat and B.M. Cohen and M.L. Evans and N. Lange",
year = "2001",
month = "9",
doi = "10.1093/biostatistics/2.3.351",
language = "English",
volume = "2",
pages = "351--364",
journal = "Biostatistics",
issn = "1465-4644",
number = "3",

}

Young, D, Ma, J, Cherkerzian, S, Froimowitz, MP, Ennulat, DJ, Cohen, BM, Evans, ML & Lange, N 2001, 'Automated identification of Fos expression' Biostatistics, vol. 2, no. 3, pp. 351-364. https://doi.org/10.1093/biostatistics/2.3.351

Automated identification of Fos expression. / Young, D.; Ma, J.; Cherkerzian, S.; Froimowitz, M.P.; Ennulat, D.J.; Cohen, B.M.; Evans, M.L.; Lange, N.

In: Biostatistics, Vol. 2, No. 3, 09.2001, p. 351-364.

Research output: Contribution to journalArticle

TY - JOUR

T1 - Automated identification of Fos expression

AU - Young, D.

AU - Ma, J.

AU - Cherkerzian, S.

AU - Froimowitz, M.P.

AU - Ennulat, D.J.

AU - Cohen, B.M.

AU - Evans, M.L.

AU - Lange, N.

PY - 2001/9

Y1 - 2001/9

N2 - The concentration of Fos, a protein encoded by the immediate-early gene c-fos, provides a measure of synaptic activity that may not parallel the electrical activity of neurons. Such a measure is important for the difficult problem of identifying dynamic properties of neuronal circuitries activated by a variety of stimuli and behaviours. We employ two-stage statistical pattern recognition to identify cellular nuclei that express Fos in two-dimensional sections of rat forebrain after administration of antipsychotic drugs. In stage one, we distinguish dark-stained candidate nuclei from image background by a thresholding algorithm and record size and shape measurements of these objects. In stage two, we compare performance of linear and quadratic discriminants, nearest-neighbour and artificial neural network classifiers that employ functions of these measurements to label candidate objects as either Fos nuclei, two touching Fos nuclei or irrelevant background material. New images of neighbouring brain tissue serve as test sets to assess generalizability of the best derived classification rule, as determined by lowest cross-validation misclassification rate. Three experts, two internal and one external, compare manual and automated results for accuracy assessment. Analyses of a subset of images on two separate occasions provide quantitative measures of inter- and intra-expert consistency. We conclude that our automated procedure yields results that compare favourably with those of the experts and thus has potential to remove much of the tedium, subjectivity and irreproducibility of current Fos identification methods in digital microscopy.

AB - The concentration of Fos, a protein encoded by the immediate-early gene c-fos, provides a measure of synaptic activity that may not parallel the electrical activity of neurons. Such a measure is important for the difficult problem of identifying dynamic properties of neuronal circuitries activated by a variety of stimuli and behaviours. We employ two-stage statistical pattern recognition to identify cellular nuclei that express Fos in two-dimensional sections of rat forebrain after administration of antipsychotic drugs. In stage one, we distinguish dark-stained candidate nuclei from image background by a thresholding algorithm and record size and shape measurements of these objects. In stage two, we compare performance of linear and quadratic discriminants, nearest-neighbour and artificial neural network classifiers that employ functions of these measurements to label candidate objects as either Fos nuclei, two touching Fos nuclei or irrelevant background material. New images of neighbouring brain tissue serve as test sets to assess generalizability of the best derived classification rule, as determined by lowest cross-validation misclassification rate. Three experts, two internal and one external, compare manual and automated results for accuracy assessment. Analyses of a subset of images on two separate occasions provide quantitative measures of inter- and intra-expert consistency. We conclude that our automated procedure yields results that compare favourably with those of the experts and thus has potential to remove much of the tedium, subjectivity and irreproducibility of current Fos identification methods in digital microscopy.

KW - Amygdala

KW - artificial neural networks

KW - digital image analysis

KW - immediate-early gene protein

KW - immunohistochemistry

KW - microscopy

KW - digital microscopy

U2 - 10.1093/biostatistics/2.3.351

DO - 10.1093/biostatistics/2.3.351

M3 - Article

VL - 2

SP - 351

EP - 364

JO - Biostatistics

T2 - Biostatistics

JF - Biostatistics

SN - 1465-4644

IS - 3

ER -

Young D, Ma J, Cherkerzian S, Froimowitz MP, Ennulat DJ, Cohen BM et al. Automated identification of Fos expression. Biostatistics. 2001 Sep;2(3):351-364. https://doi.org/10.1093/biostatistics/2.3.351