Hotelling trace criterion and its correlation with human-observer performance.

R. D. Fiete, Harrison H Barrett, W. E. Smith, K. J. Myers

Research output: Contribution to journalArticle

127 Citations (Scopus)

Abstract

The Hotelling trace criterion (HTC) is used to find a set of linear features that optimally separate two classes of objects. The objects used in our study were simulated livers with and without tumors, with noise, blur, and object variability. Using the receiver-operating-characteristic parameter da as our measure, we have found that the ability of the HTC to separate these objects into their correct classes, by detecting the presence or absence of a tumor, has a correlation of 0.988 with the ability of humans to separate the same two classes of objects. This suggests, therefore, that the HTC can be used as a figure of merit for optimizing system parameters, since it calculates a single, scalar figure of merit that has a high correlation with human-observer performance.

Original languageEnglish (US)
Pages (from-to)945-953
Number of pages9
JournalJournal of the Optical Society of America. A, Optics and image science
Volume4
Issue number5
StatePublished - May 1987

Fingerprint

Tumors
ROC Curve
Liver
Noise
Neoplasms

ASJC Scopus subject areas

  • Medicine(all)

Cite this

Hotelling trace criterion and its correlation with human-observer performance. / Fiete, R. D.; Barrett, Harrison H; Smith, W. E.; Myers, K. J.

In: Journal of the Optical Society of America. A, Optics and image science, Vol. 4, No. 5, 05.1987, p. 945-953.

Research output: Contribution to journalArticle

@article{2965a5d7303745deb7ffc88c0dcd8bb9,
title = "Hotelling trace criterion and its correlation with human-observer performance.",
abstract = "The Hotelling trace criterion (HTC) is used to find a set of linear features that optimally separate two classes of objects. The objects used in our study were simulated livers with and without tumors, with noise, blur, and object variability. Using the receiver-operating-characteristic parameter da as our measure, we have found that the ability of the HTC to separate these objects into their correct classes, by detecting the presence or absence of a tumor, has a correlation of 0.988 with the ability of humans to separate the same two classes of objects. This suggests, therefore, that the HTC can be used as a figure of merit for optimizing system parameters, since it calculates a single, scalar figure of merit that has a high correlation with human-observer performance.",
author = "Fiete, {R. D.} and Barrett, {Harrison H} and Smith, {W. E.} and Myers, {K. J.}",
year = "1987",
month = "5",
language = "English (US)",
volume = "4",
pages = "945--953",
journal = "Journal of the Optical Society of America A: Optics and Image Science, and Vision",
issn = "1084-7529",
publisher = "The Optical Society",
number = "5",

}

TY - JOUR

T1 - Hotelling trace criterion and its correlation with human-observer performance.

AU - Fiete, R. D.

AU - Barrett, Harrison H

AU - Smith, W. E.

AU - Myers, K. J.

PY - 1987/5

Y1 - 1987/5

N2 - The Hotelling trace criterion (HTC) is used to find a set of linear features that optimally separate two classes of objects. The objects used in our study were simulated livers with and without tumors, with noise, blur, and object variability. Using the receiver-operating-characteristic parameter da as our measure, we have found that the ability of the HTC to separate these objects into their correct classes, by detecting the presence or absence of a tumor, has a correlation of 0.988 with the ability of humans to separate the same two classes of objects. This suggests, therefore, that the HTC can be used as a figure of merit for optimizing system parameters, since it calculates a single, scalar figure of merit that has a high correlation with human-observer performance.

AB - The Hotelling trace criterion (HTC) is used to find a set of linear features that optimally separate two classes of objects. The objects used in our study were simulated livers with and without tumors, with noise, blur, and object variability. Using the receiver-operating-characteristic parameter da as our measure, we have found that the ability of the HTC to separate these objects into their correct classes, by detecting the presence or absence of a tumor, has a correlation of 0.988 with the ability of humans to separate the same two classes of objects. This suggests, therefore, that the HTC can be used as a figure of merit for optimizing system parameters, since it calculates a single, scalar figure of merit that has a high correlation with human-observer performance.

UR - http://www.scopus.com/inward/record.url?scp=0023335011&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0023335011&partnerID=8YFLogxK

M3 - Article

C2 - 3598746

AN - SCOPUS:0023335011

VL - 4

SP - 945

EP - 953

JO - Journal of the Optical Society of America A: Optics and Image Science, and Vision

JF - Journal of the Optical Society of America A: Optics and Image Science, and Vision

SN - 1084-7529

IS - 5

ER -