Method for counting mitoses by image processing in Feulgen stained breast cancer sections

T K ten Kate, J A Beliën, A W Smeulders, J P Baak

Research output: Contribution to journalArticleAcademicpeer-review

Abstract

This study describes an image processing method for the assessment of the mitotic count in Feulgen-stained breast cancer sections. The segmentation procedure was optimized to eliminate 95-98% of the nonmitoses, whereas 11% of the mitoses did not survive the segmentation procedure. Contour features and optical density measurements of the remaining objects were computed to allow for classification. Twelve specimens were analyzed, nine used to serve as a training set, and three put aside for later use as independent test set. The fully automatic image processing method correctly classified 81% of the mitoses at the specimen level while inserting 30% false positives. The automatic procedure strongly correlated with the interactive counting procedure (r = 0.98). Although the fully automatic method provided satisfactory results, it is not yet suited for clinical practice. The automated method with an interactive evaluation step gave an accurate reflection of the mitotic count showing an almost perfect correlation with the results of the interactive morphometry (r = 0.998). Therefore this semiautomated method may be useful as prescreening device.

Original languageEnglish
Pages (from-to)241-50
Number of pages10
JournalCytometry Part B. Clinical Cytometry
Volume14
Issue number3
DOIs
Publication statusPublished - 1993

Cite this

@article{669d0ed300d2443d9313562d67cb3b05,
title = "Method for counting mitoses by image processing in Feulgen stained breast cancer sections",
abstract = "This study describes an image processing method for the assessment of the mitotic count in Feulgen-stained breast cancer sections. The segmentation procedure was optimized to eliminate 95-98{\%} of the nonmitoses, whereas 11{\%} of the mitoses did not survive the segmentation procedure. Contour features and optical density measurements of the remaining objects were computed to allow for classification. Twelve specimens were analyzed, nine used to serve as a training set, and three put aside for later use as independent test set. The fully automatic image processing method correctly classified 81{\%} of the mitoses at the specimen level while inserting 30{\%} false positives. The automatic procedure strongly correlated with the interactive counting procedure (r = 0.98). Although the fully automatic method provided satisfactory results, it is not yet suited for clinical practice. The automated method with an interactive evaluation step gave an accurate reflection of the mitotic count showing an almost perfect correlation with the results of the interactive morphometry (r = 0.998). Therefore this semiautomated method may be useful as prescreening device.",
keywords = "Automation, Breast Neoplasms, Cell Count, Coloring Agents, False Positive Reactions, Humans, Image Processing, Computer-Assisted, Mitosis, Rosaniline Dyes, Software, Journal Article, Research Support, Non-U.S. Gov't",
author = "{ten Kate}, {T K} and Beli{\"e}n, {J A} and Smeulders, {A W} and Baak, {J P}",
year = "1993",
doi = "10.1002/cyto.990140302",
language = "English",
volume = "14",
pages = "241--50",
journal = "Cytometry Part B. Clinical Cytometry",
issn = "1552-4949",
publisher = "Wiley-Liss Inc.",
number = "3",

}

Method for counting mitoses by image processing in Feulgen stained breast cancer sections. / ten Kate, T K; Beliën, J A; Smeulders, A W; Baak, J P.

In: Cytometry Part B. Clinical Cytometry, Vol. 14, No. 3, 1993, p. 241-50.

Research output: Contribution to journalArticleAcademicpeer-review

TY - JOUR

T1 - Method for counting mitoses by image processing in Feulgen stained breast cancer sections

AU - ten Kate, T K

AU - Beliën, J A

AU - Smeulders, A W

AU - Baak, J P

PY - 1993

Y1 - 1993

N2 - This study describes an image processing method for the assessment of the mitotic count in Feulgen-stained breast cancer sections. The segmentation procedure was optimized to eliminate 95-98% of the nonmitoses, whereas 11% of the mitoses did not survive the segmentation procedure. Contour features and optical density measurements of the remaining objects were computed to allow for classification. Twelve specimens were analyzed, nine used to serve as a training set, and three put aside for later use as independent test set. The fully automatic image processing method correctly classified 81% of the mitoses at the specimen level while inserting 30% false positives. The automatic procedure strongly correlated with the interactive counting procedure (r = 0.98). Although the fully automatic method provided satisfactory results, it is not yet suited for clinical practice. The automated method with an interactive evaluation step gave an accurate reflection of the mitotic count showing an almost perfect correlation with the results of the interactive morphometry (r = 0.998). Therefore this semiautomated method may be useful as prescreening device.

AB - This study describes an image processing method for the assessment of the mitotic count in Feulgen-stained breast cancer sections. The segmentation procedure was optimized to eliminate 95-98% of the nonmitoses, whereas 11% of the mitoses did not survive the segmentation procedure. Contour features and optical density measurements of the remaining objects were computed to allow for classification. Twelve specimens were analyzed, nine used to serve as a training set, and three put aside for later use as independent test set. The fully automatic image processing method correctly classified 81% of the mitoses at the specimen level while inserting 30% false positives. The automatic procedure strongly correlated with the interactive counting procedure (r = 0.98). Although the fully automatic method provided satisfactory results, it is not yet suited for clinical practice. The automated method with an interactive evaluation step gave an accurate reflection of the mitotic count showing an almost perfect correlation with the results of the interactive morphometry (r = 0.998). Therefore this semiautomated method may be useful as prescreening device.

KW - Automation

KW - Breast Neoplasms

KW - Cell Count

KW - Coloring Agents

KW - False Positive Reactions

KW - Humans

KW - Image Processing, Computer-Assisted

KW - Mitosis

KW - Rosaniline Dyes

KW - Software

KW - Journal Article

KW - Research Support, Non-U.S. Gov't

U2 - 10.1002/cyto.990140302

DO - 10.1002/cyto.990140302

M3 - Article

VL - 14

SP - 241

EP - 250

JO - Cytometry Part B. Clinical Cytometry

JF - Cytometry Part B. Clinical Cytometry

SN - 1552-4949

IS - 3

ER -