Real Data Augmentation for Medical Image Classification

Chuanhai Zhang; Wallapak Tavanapong; Johnny Wong; Piet C. de Groen; Jung Hwan Oh

doi:10.1007/978-3-319-67534-3_8

Real Data Augmentation for Medical Image Classification

Chuanhai Zhang, Wallapak Tavanapong, Johnny Wong, Piet C. de Groen, Jung Hwan Oh

Medicine - Gastro, Hepatology, Nutrition Division

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

18 Scopus citations

Abstract

Many medical image classification tasks share a common unbalanced data problem. That is images of the target classes, e.g., certain types of diseases, only appear in a very small portion of the entire dataset. Nowadays, large collections of medical images are readily available. However, it is costly and may not even be feasible for medical experts to manually comb through a huge unlabeled dataset to obtain enough representative examples of the rare classes. In this paper, we propose a new method called Unified LF&SM to recommend most similar images for each class from a large unlabeled dataset for verification by medical experts and inclusion in the seed labeled dataset. Our real data augmentation significantly reduces expensive manual labeling time. In our experiments, Unified LF&SM performed best, selecting a high percentage of relevant images in its recommendation and achieving the best classification accuracy. It is easily extendable to other medical image classification problems.

Original language	English (US)
Title of host publication	Intravascular Imaging and Computer Assisted Stenting, and Large-Scale Annotation of Biomedical Data and Expert Label Synthesis - 6th Joint International Workshops, CVII-STENT 2017 and 2nd International Workshop, LABELS 2017 Held in Conjunction with MICCAI 2017, Proceedings
Editors	Tal Arbel, M. Jorge Cardoso
Publisher	Springer Verlag
Pages	67-76
Number of pages	10
ISBN (Print)	9783319675336
DOIs	https://doi.org/10.1007/978-3-319-67534-3_8
State	Published - 2017
Event	6th Joint International Workshops on Computing and Visualization for Intravascular Imaging and Computer Assisted Stenting, CVII-STENT 2017 and 2nd International Workshop on Large-Scale Annotation of Biomedical Data and Expert Label Synthesis, LABELS 2017 held in Conjunction with 20th International Conference on Medical Image Computing and Computer Assisted Intervention, MICCAI 2017 - Quebec City, Canada Duration: Sep 10 2017 → Sep 14 2017

Publication series

Name	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume	10552 LNCS
ISSN (Print)	0302-9743
ISSN (Electronic)	1611-3349

Other

Other	6th Joint International Workshops on Computing and Visualization for Intravascular Imaging and Computer Assisted Stenting, CVII-STENT 2017 and 2nd International Workshop on Large-Scale Annotation of Biomedical Data and Expert Label Synthesis, LABELS 2017 held in Conjunction with 20th International Conference on Medical Image Computing and Computer Assisted Intervention, MICCAI 2017
Country/Territory	Canada
City	Quebec City
Period	9/10/17 → 9/14/17

Bibliographical note

Publisher Copyright:
© 2017, Springer International Publishing AG.

Keywords

Image classification
Real data augmentation
Unbalanced data

Publisher link

10.1007/978-3-319-67534-3_8

Find It

Find It at U of M Libraries

Cite this

Zhang, C., Tavanapong, W., Wong, J., de Groen, P. C., & Oh, J. H. (2017). Real Data Augmentation for Medical Image Classification. In T. Arbel, & M. J. Cardoso (Eds.), Intravascular Imaging and Computer Assisted Stenting, and Large-Scale Annotation of Biomedical Data and Expert Label Synthesis - 6th Joint International Workshops, CVII-STENT 2017 and 2nd International Workshop, LABELS 2017 Held in Conjunction with MICCAI 2017, Proceedings (pp. 67-76). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 10552 LNCS). Springer Verlag. https://doi.org/10.1007/978-3-319-67534-3_8

Real Data Augmentation for Medical Image Classification. / Zhang, Chuanhai; Tavanapong, Wallapak; Wong, Johnny et al.
Intravascular Imaging and Computer Assisted Stenting, and Large-Scale Annotation of Biomedical Data and Expert Label Synthesis - 6th Joint International Workshops, CVII-STENT 2017 and 2nd International Workshop, LABELS 2017 Held in Conjunction with MICCAI 2017, Proceedings. ed. / Tal Arbel; M. Jorge Cardoso. Springer Verlag, 2017. p. 67-76 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 10552 LNCS).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

Zhang, C, Tavanapong, W, Wong, J, de Groen, PC & Oh, JH 2017, Real Data Augmentation for Medical Image Classification. in T Arbel & MJ Cardoso (eds), Intravascular Imaging and Computer Assisted Stenting, and Large-Scale Annotation of Biomedical Data and Expert Label Synthesis - 6th Joint International Workshops, CVII-STENT 2017 and 2nd International Workshop, LABELS 2017 Held in Conjunction with MICCAI 2017, Proceedings. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 10552 LNCS, Springer Verlag, pp. 67-76, 6th Joint International Workshops on Computing and Visualization for Intravascular Imaging and Computer Assisted Stenting, CVII-STENT 2017 and 2nd International Workshop on Large-Scale Annotation of Biomedical Data and Expert Label Synthesis, LABELS 2017 held in Conjunction with 20th International Conference on Medical Image Computing and Computer Assisted Intervention, MICCAI 2017, Quebec City, Canada, 9/10/17. https://doi.org/10.1007/978-3-319-67534-3_8

Zhang C, Tavanapong W, Wong J, de Groen PC, Oh JH. Real Data Augmentation for Medical Image Classification. In Arbel T, Cardoso MJ, editors, Intravascular Imaging and Computer Assisted Stenting, and Large-Scale Annotation of Biomedical Data and Expert Label Synthesis - 6th Joint International Workshops, CVII-STENT 2017 and 2nd International Workshop, LABELS 2017 Held in Conjunction with MICCAI 2017, Proceedings. Springer Verlag. 2017. p. 67-76. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). doi: 10.1007/978-3-319-67534-3_8

Zhang, Chuanhai ; Tavanapong, Wallapak ; Wong, Johnny et al. / Real Data Augmentation for Medical Image Classification. Intravascular Imaging and Computer Assisted Stenting, and Large-Scale Annotation of Biomedical Data and Expert Label Synthesis - 6th Joint International Workshops, CVII-STENT 2017 and 2nd International Workshop, LABELS 2017 Held in Conjunction with MICCAI 2017, Proceedings. editor / Tal Arbel ; M. Jorge Cardoso. Springer Verlag, 2017. pp. 67-76 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).

@inproceedings{86014e4f11b14cb899c5f3d8fdfb1d74,

title = "Real Data Augmentation for Medical Image Classification",

abstract = "Many medical image classification tasks share a common unbalanced data problem. That is images of the target classes, e.g., certain types of diseases, only appear in a very small portion of the entire dataset. Nowadays, large collections of medical images are readily available. However, it is costly and may not even be feasible for medical experts to manually comb through a huge unlabeled dataset to obtain enough representative examples of the rare classes. In this paper, we propose a new method called Unified LF&SM to recommend most similar images for each class from a large unlabeled dataset for verification by medical experts and inclusion in the seed labeled dataset. Our real data augmentation significantly reduces expensive manual labeling time. In our experiments, Unified LF&SM performed best, selecting a high percentage of relevant images in its recommendation and achieving the best classification accuracy. It is easily extendable to other medical image classification problems.",

keywords = "Image classification, Real data augmentation, Unbalanced data",

author = "Chuanhai Zhang and Wallapak Tavanapong and Johnny Wong and {de Groen}, {Piet C.} and Oh, {Jung Hwan}",

note = "Publisher Copyright: {\textcopyright} 2017, Springer International Publishing AG.; 6th Joint International Workshops on Computing and Visualization for Intravascular Imaging and Computer Assisted Stenting, CVII-STENT 2017 and 2nd International Workshop on Large-Scale Annotation of Biomedical Data and Expert Label Synthesis, LABELS 2017 held in Conjunction with 20th International Conference on Medical Image Computing and Computer Assisted Intervention, MICCAI 2017 ; Conference date: 10-09-2017 Through 14-09-2017",

year = "2017",

doi = "10.1007/978-3-319-67534-3_8",

language = "English (US)",

isbn = "9783319675336",

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

publisher = "Springer Verlag",

pages = "67--76",

editor = "Tal Arbel and Cardoso, {M. Jorge}",

booktitle = "Intravascular Imaging and Computer Assisted Stenting, and Large-Scale Annotation of Biomedical Data and Expert Label Synthesis - 6th Joint International Workshops, CVII-STENT 2017 and 2nd International Workshop, LABELS 2017 Held in Conjunction with MICCAI 2017, Proceedings",

}

TY - GEN

T1 - Real Data Augmentation for Medical Image Classification

AU - Zhang, Chuanhai

AU - Tavanapong, Wallapak

AU - Wong, Johnny

AU - de Groen, Piet C.

AU - Oh, Jung Hwan

PY - 2017

Y1 - 2017

N2 - Many medical image classification tasks share a common unbalanced data problem. That is images of the target classes, e.g., certain types of diseases, only appear in a very small portion of the entire dataset. Nowadays, large collections of medical images are readily available. However, it is costly and may not even be feasible for medical experts to manually comb through a huge unlabeled dataset to obtain enough representative examples of the rare classes. In this paper, we propose a new method called Unified LF&SM to recommend most similar images for each class from a large unlabeled dataset for verification by medical experts and inclusion in the seed labeled dataset. Our real data augmentation significantly reduces expensive manual labeling time. In our experiments, Unified LF&SM performed best, selecting a high percentage of relevant images in its recommendation and achieving the best classification accuracy. It is easily extendable to other medical image classification problems.

AB - Many medical image classification tasks share a common unbalanced data problem. That is images of the target classes, e.g., certain types of diseases, only appear in a very small portion of the entire dataset. Nowadays, large collections of medical images are readily available. However, it is costly and may not even be feasible for medical experts to manually comb through a huge unlabeled dataset to obtain enough representative examples of the rare classes. In this paper, we propose a new method called Unified LF&SM to recommend most similar images for each class from a large unlabeled dataset for verification by medical experts and inclusion in the seed labeled dataset. Our real data augmentation significantly reduces expensive manual labeling time. In our experiments, Unified LF&SM performed best, selecting a high percentage of relevant images in its recommendation and achieving the best classification accuracy. It is easily extendable to other medical image classification problems.

KW - Image classification

KW - Real data augmentation

KW - Unbalanced data

UR - http://www.scopus.com/inward/record.url?scp=85029805460&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85029805460&partnerID=8YFLogxK

U2 - 10.1007/978-3-319-67534-3_8

DO - 10.1007/978-3-319-67534-3_8

M3 - Conference contribution

AN - SCOPUS:85029805460

SN - 9783319675336

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

SP - 67

EP - 76

BT - Intravascular Imaging and Computer Assisted Stenting, and Large-Scale Annotation of Biomedical Data and Expert Label Synthesis - 6th Joint International Workshops, CVII-STENT 2017 and 2nd International Workshop, LABELS 2017 Held in Conjunction with MICCAI 2017, Proceedings

A2 - Arbel, Tal

A2 - Cardoso, M. Jorge

PB - Springer Verlag

T2 - 6th Joint International Workshops on Computing and Visualization for Intravascular Imaging and Computer Assisted Stenting, CVII-STENT 2017 and 2nd International Workshop on Large-Scale Annotation of Biomedical Data and Expert Label Synthesis, LABELS 2017 held in Conjunction with 20th International Conference on Medical Image Computing and Computer Assisted Intervention, MICCAI 2017

Y2 - 10 September 2017 through 14 September 2017

ER -

Real Data Augmentation for Medical Image Classification

Abstract

Publication series

Other

Bibliographical note

Keywords

Publisher link

Find It

Other files and links

Fingerprint

Cite this