Abstract

The use of high-resolution digital images of histopathology slides as a routine diagnostic tool for surgical pathology was investigated. The study purpose was to determine the diagnostic concordance between pathologic interpretations using whole-slide imaging and standard light microscopy. Two hundred fifty-one consecutive surgical pathology cases (312 parts, 1085 slides) from a single pathology service were included in the study after cases had been signed out and reports generated. A broad array of diagnostic challenges and tissue sources were represented, including 52 neoplastic cases. All cases were digitized at ×20 and presented to 2 pathologists for diagnosis using whole-slide imaging as the sole diagnostic tool. Diagnoses rendered by the whole-slide imaging pathologists were compared with the original light microscopy diagnoses. Overall concordance between whole-slide imaging and light microscopy as determined by a third pathologist and jury panel was 96.5% (95% confidence interval, 94.8%-98.3%). Concordance between whole-slide imaging pathologists was 97.7% (95% confidence interval, 94.7%-99.2%). Five cases were discordant between the whole-slide imaging diagnosis and the original light microscopy diagnosis, of which 2 were clinically significant. Discordance resulted from interpretive criteria or diagnostic error. The whole-slide imaging modality did not contribute to diagnostic differences. Problems encountered by the whole-slide imaging pathologists primarily involved the inability to clearly visualize nuclear detail or microscopic organisms. Technical difficulties associated with image scanning required at least 1 slide be rescanned in 13% of the cases. Technical and operational issues associated with whole-slide imaging scanning devices used in this study were found to be the most significant obstacle to the use of whole-slide imaging in general surgical pathology.

Original languageEnglish (US)
Pages (from-to)1739-1744
Number of pages6
JournalHuman Pathology
Volume43
Issue number10
DOIs
StatePublished - Oct 1 2012

Fingerprint

Surgical Pathology
Microscopy
Light
Confidence Intervals
Diagnostic Errors
Pathologists
Pathology
Equipment and Supplies

Keywords

  • Telemedicine
  • Virtual microscopy
  • Whole-slide imaging

ASJC Scopus subject areas

  • Pathology and Forensic Medicine

Cite this

@article{933283a7a19a4a4db7874c646aaf4f92,
title = "Concordance between whole-slide imaging and light microscopy for routine surgical pathology",
abstract = "The use of high-resolution digital images of histopathology slides as a routine diagnostic tool for surgical pathology was investigated. The study purpose was to determine the diagnostic concordance between pathologic interpretations using whole-slide imaging and standard light microscopy. Two hundred fifty-one consecutive surgical pathology cases (312 parts, 1085 slides) from a single pathology service were included in the study after cases had been signed out and reports generated. A broad array of diagnostic challenges and tissue sources were represented, including 52 neoplastic cases. All cases were digitized at ×20 and presented to 2 pathologists for diagnosis using whole-slide imaging as the sole diagnostic tool. Diagnoses rendered by the whole-slide imaging pathologists were compared with the original light microscopy diagnoses. Overall concordance between whole-slide imaging and light microscopy as determined by a third pathologist and jury panel was 96.5{\%} (95{\%} confidence interval, 94.8{\%}-98.3{\%}). Concordance between whole-slide imaging pathologists was 97.7{\%} (95{\%} confidence interval, 94.7{\%}-99.2{\%}). Five cases were discordant between the whole-slide imaging diagnosis and the original light microscopy diagnosis, of which 2 were clinically significant. Discordance resulted from interpretive criteria or diagnostic error. The whole-slide imaging modality did not contribute to diagnostic differences. Problems encountered by the whole-slide imaging pathologists primarily involved the inability to clearly visualize nuclear detail or microscopic organisms. Technical difficulties associated with image scanning required at least 1 slide be rescanned in 13{\%} of the cases. Technical and operational issues associated with whole-slide imaging scanning devices used in this study were found to be the most significant obstacle to the use of whole-slide imaging in general surgical pathology.",
keywords = "Telemedicine, Virtual microscopy, Whole-slide imaging",
author = "Campbell, {Walter S} and Lele, {Subodh M} and West, {William W.} and Lazenby, {Audrey J} and Smith, {Lynette M} and Hinrichs, {Steven Heye}",
year = "2012",
month = "10",
day = "1",
doi = "10.1016/j.humpath.2011.12.023",
language = "English (US)",
volume = "43",
pages = "1739--1744",
journal = "Human Pathology",
issn = "0046-8177",
publisher = "W.B. Saunders Ltd",
number = "10",

}

TY - JOUR

T1 - Concordance between whole-slide imaging and light microscopy for routine surgical pathology

AU - Campbell, Walter S

AU - Lele, Subodh M

AU - West, William W.

AU - Lazenby, Audrey J

AU - Smith, Lynette M

AU - Hinrichs, Steven Heye

PY - 2012/10/1

Y1 - 2012/10/1

N2 - The use of high-resolution digital images of histopathology slides as a routine diagnostic tool for surgical pathology was investigated. The study purpose was to determine the diagnostic concordance between pathologic interpretations using whole-slide imaging and standard light microscopy. Two hundred fifty-one consecutive surgical pathology cases (312 parts, 1085 slides) from a single pathology service were included in the study after cases had been signed out and reports generated. A broad array of diagnostic challenges and tissue sources were represented, including 52 neoplastic cases. All cases were digitized at ×20 and presented to 2 pathologists for diagnosis using whole-slide imaging as the sole diagnostic tool. Diagnoses rendered by the whole-slide imaging pathologists were compared with the original light microscopy diagnoses. Overall concordance between whole-slide imaging and light microscopy as determined by a third pathologist and jury panel was 96.5% (95% confidence interval, 94.8%-98.3%). Concordance between whole-slide imaging pathologists was 97.7% (95% confidence interval, 94.7%-99.2%). Five cases were discordant between the whole-slide imaging diagnosis and the original light microscopy diagnosis, of which 2 were clinically significant. Discordance resulted from interpretive criteria or diagnostic error. The whole-slide imaging modality did not contribute to diagnostic differences. Problems encountered by the whole-slide imaging pathologists primarily involved the inability to clearly visualize nuclear detail or microscopic organisms. Technical difficulties associated with image scanning required at least 1 slide be rescanned in 13% of the cases. Technical and operational issues associated with whole-slide imaging scanning devices used in this study were found to be the most significant obstacle to the use of whole-slide imaging in general surgical pathology.

AB - The use of high-resolution digital images of histopathology slides as a routine diagnostic tool for surgical pathology was investigated. The study purpose was to determine the diagnostic concordance between pathologic interpretations using whole-slide imaging and standard light microscopy. Two hundred fifty-one consecutive surgical pathology cases (312 parts, 1085 slides) from a single pathology service were included in the study after cases had been signed out and reports generated. A broad array of diagnostic challenges and tissue sources were represented, including 52 neoplastic cases. All cases were digitized at ×20 and presented to 2 pathologists for diagnosis using whole-slide imaging as the sole diagnostic tool. Diagnoses rendered by the whole-slide imaging pathologists were compared with the original light microscopy diagnoses. Overall concordance between whole-slide imaging and light microscopy as determined by a third pathologist and jury panel was 96.5% (95% confidence interval, 94.8%-98.3%). Concordance between whole-slide imaging pathologists was 97.7% (95% confidence interval, 94.7%-99.2%). Five cases were discordant between the whole-slide imaging diagnosis and the original light microscopy diagnosis, of which 2 were clinically significant. Discordance resulted from interpretive criteria or diagnostic error. The whole-slide imaging modality did not contribute to diagnostic differences. Problems encountered by the whole-slide imaging pathologists primarily involved the inability to clearly visualize nuclear detail or microscopic organisms. Technical difficulties associated with image scanning required at least 1 slide be rescanned in 13% of the cases. Technical and operational issues associated with whole-slide imaging scanning devices used in this study were found to be the most significant obstacle to the use of whole-slide imaging in general surgical pathology.

KW - Telemedicine

KW - Virtual microscopy

KW - Whole-slide imaging

UR - http://www.scopus.com/inward/record.url?scp=84866492842&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84866492842&partnerID=8YFLogxK

U2 - 10.1016/j.humpath.2011.12.023

DO - 10.1016/j.humpath.2011.12.023

M3 - Article

VL - 43

SP - 1739

EP - 1744

JO - Human Pathology

JF - Human Pathology

SN - 0046-8177

IS - 10

ER -