A comparison between a deep convolutional neural network and radiologists for classifying regions of interest in mammography

Thijs Kooi; Albert Gubern-Merida; Jan Jurre Mordang; Ritse M. Mann; Ruud Pijnappel; Klaas H. Schuur; Ard den Heeten; Nico Karssemeijer

doi:10.1007/978-3-319-41546-8_7

A comparison between a deep convolutional neural network and radiologists for classifying regions of interest in mammography

Thijs Kooi^*, Albert Gubern-Merida, Jan Jurre Mordang, Ritse M. Mann, Ruud Pijnappel, Klaas H. Schuur, Ard den Heeten, Nico Karssemeijer

^*Corresponding author for this work

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › Academic › peer-review

3 Citations (Scopus)

Abstract

In this paper, we employ a deep Convolutional Neural Network (CNN) for the classification of regions of interest of malignant soft tissue lesions in mammography and show that it performs on par to experienced radiologists. The CNN was applied to 398 regions of 5×5 cm, half of which contained a malignant lesion and the other half depicted suspicious regions in normal mammograms detected by a traditional CAD system. Four radiologists participated in the study. ROC analysis was used for evaluating results. The AUC of CNN was 0.87, which was higher than the mean AUC of the radiologists (0.84), though the difference was not significant.

Original language	English
Title of host publication	Breast Imaging - 13th International Workshop, IWDM 2016, Proceedings
Publisher	Springer-Verlag
Pages	51-56
Number of pages	6
Volume	9699
ISBN (Print)	9783319415451
DOIs	https://doi.org/10.1007/978-3-319-41546-8_7
Publication status	Published - 2016
Externally published	Yes
Event	13th International Workshop on Breast Imaging, IWDM 2016 - Malmo, Sweden Duration: 19 Jun 2016 → 22 Jun 2016

Publication series

Name	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume	9699
ISSN (Print)	03029743
ISSN (Electronic)	16113349

Conference

Conference	13th International Workshop on Breast Imaging, IWDM 2016
Country/Territory	Sweden
City	Malmo
Period	19/06/16 → 22/06/16

Access to Document

10.1007/978-3-319-41546-8_7

Cite this

Kooi, T., Gubern-Merida, A., Mordang, J. J., Mann, R. M., Pijnappel, R., Schuur, K. H., den Heeten, A., & Karssemeijer, N. (2016). A comparison between a deep convolutional neural network and radiologists for classifying regions of interest in mammography. In Breast Imaging - 13th International Workshop, IWDM 2016, Proceedings (Vol. 9699, pp. 51-56). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 9699). Springer-Verlag. https://doi.org/10.1007/978-3-319-41546-8_7

Kooi, Thijs ; Gubern-Merida, Albert ; Mordang, Jan Jurre et al. / A comparison between a deep convolutional neural network and radiologists for classifying regions of interest in mammography. Breast Imaging - 13th International Workshop, IWDM 2016, Proceedings. Vol. 9699 Springer-Verlag, 2016. pp. 51-56 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).

@inproceedings{15e42cf1a36944c289d9fd2780eb9cba,

title = "A comparison between a deep convolutional neural network and radiologists for classifying regions of interest in mammography",

abstract = "In this paper, we employ a deep Convolutional Neural Network (CNN) for the classification of regions of interest of malignant soft tissue lesions in mammography and show that it performs on par to experienced radiologists. The CNN was applied to 398 regions of 5×5 cm, half of which contained a malignant lesion and the other half depicted suspicious regions in normal mammograms detected by a traditional CAD system. Four radiologists participated in the study. ROC analysis was used for evaluating results. The AUC of CNN was 0.87, which was higher than the mean AUC of the radiologists (0.84), though the difference was not significant.",

author = "Thijs Kooi and Albert Gubern-Merida and Mordang, \{Jan Jurre\} and Mann, \{Ritse M.\} and Ruud Pijnappel and Schuur, \{Klaas H.\} and \{den Heeten\}, Ard and Nico Karssemeijer",

year = "2016",

doi = "10.1007/978-3-319-41546-8\_7",

language = "English",

isbn = "9783319415451",

volume = "9699",

series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",

publisher = "Springer-Verlag",

pages = "51--56",

booktitle = "Breast Imaging - 13th International Workshop, IWDM 2016, Proceedings",

address = "Germany",

note = "13th International Workshop on Breast Imaging, IWDM 2016 ; Conference date: 19-06-2016 Through 22-06-2016",

}

Kooi, T, Gubern-Merida, A, Mordang, JJ, Mann, RM, Pijnappel, R, Schuur, KH, den Heeten, A & Karssemeijer, N 2016, A comparison between a deep convolutional neural network and radiologists for classifying regions of interest in mammography. in Breast Imaging - 13th International Workshop, IWDM 2016, Proceedings. vol. 9699, Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 9699, Springer-Verlag, pp. 51-56, 13th International Workshop on Breast Imaging, IWDM 2016, Malmo, Sweden, 19/06/16. https://doi.org/10.1007/978-3-319-41546-8_7

A comparison between a deep convolutional neural network and radiologists for classifying regions of interest in mammography. / Kooi, Thijs; Gubern-Merida, Albert; Mordang, Jan Jurre et al.
Breast Imaging - 13th International Workshop, IWDM 2016, Proceedings. Vol. 9699 Springer-Verlag, 2016. p. 51-56 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 9699).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › Academic › peer-review

TY - GEN

T1 - A comparison between a deep convolutional neural network and radiologists for classifying regions of interest in mammography

AU - Kooi, Thijs

AU - Gubern-Merida, Albert

AU - Mordang, Jan Jurre

AU - Mann, Ritse M.

AU - Pijnappel, Ruud

AU - Schuur, Klaas H.

AU - den Heeten, Ard

AU - Karssemeijer, Nico

PY - 2016

Y1 - 2016

N2 - In this paper, we employ a deep Convolutional Neural Network (CNN) for the classification of regions of interest of malignant soft tissue lesions in mammography and show that it performs on par to experienced radiologists. The CNN was applied to 398 regions of 5×5 cm, half of which contained a malignant lesion and the other half depicted suspicious regions in normal mammograms detected by a traditional CAD system. Four radiologists participated in the study. ROC analysis was used for evaluating results. The AUC of CNN was 0.87, which was higher than the mean AUC of the radiologists (0.84), though the difference was not significant.

AB - In this paper, we employ a deep Convolutional Neural Network (CNN) for the classification of regions of interest of malignant soft tissue lesions in mammography and show that it performs on par to experienced radiologists. The CNN was applied to 398 regions of 5×5 cm, half of which contained a malignant lesion and the other half depicted suspicious regions in normal mammograms detected by a traditional CAD system. Four radiologists participated in the study. ROC analysis was used for evaluating results. The AUC of CNN was 0.87, which was higher than the mean AUC of the radiologists (0.84), though the difference was not significant.

UR - https://www.scopus.com/pages/publications/84977600905

U2 - 10.1007/978-3-319-41546-8_7

DO - 10.1007/978-3-319-41546-8_7

M3 - Conference contribution

AN - SCOPUS:84977600905

SN - 9783319415451

VL - 9699

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 51

EP - 56

BT - Breast Imaging - 13th International Workshop, IWDM 2016, Proceedings

PB - Springer-Verlag

T2 - 13th International Workshop on Breast Imaging, IWDM 2016

Y2 - 19 June 2016 through 22 June 2016

ER -

Kooi T, Gubern-Merida A, Mordang JJ, Mann RM, Pijnappel R, Schuur KH et al. A comparison between a deep convolutional neural network and radiologists for classifying regions of interest in mammography. In Breast Imaging - 13th International Workshop, IWDM 2016, Proceedings. Vol. 9699. Springer-Verlag. 2016. p. 51-56. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). doi: 10.1007/978-3-319-41546-8_7

A comparison between a deep convolutional neural network and radiologists for classifying regions of interest in mammography

Abstract

Publication series

Conference

Access to Document

Other files and links

Fingerprint

Cite this