Presentation 2022-09-15
Learning of Squamous Cell Image Classification Model Using Preference Learning to Assist Cervical Cytology
Yuta Nambu, Tasuku Mariya, Syota Shinkai, Mina Umemoto, Hiroko Asanuma, Yoshihiko Hirohashi, Tsuyoshi Saito, Toshihiko Torigoe, Ikuma Sato, Yuichi Fujino,
PDF Download Page PDF download Page Link
Abstract(in Japanese) (See Japanese page)
Abstract(in English) To support cervical cell diagnosis, Various classification methods of cervical cell images using machine learning have been proposed. However, even methods with large amounts of data and large models have not been able to achieve classification with sufficient performance. One reason for this difficulty may be that label noise tends to occur due to the difficulty of making decisions. Therefore, we propose preference learning of a model that discriminates order relationships among cell images using pairwise data such that `Image B is more malignant than Image A' as labels. In this paper, we report the performance of a deep learning model with preference learning on cervical cell images, and compare the discriminative performance of the model with classification learning.
Keyword(in Japanese) (See Japanese page)
Keyword(in English) Cell image classification / Learning to rank / Deep learning
Paper # MI2022-62
Date of Issue 2022-09-08 (MI)

Conference Information
Committee MI
Conference Date 2022/9/15(1days)
Place (in Japanese) (See Japanese page)
Place (in English)
Topics (in Japanese) (See Japanese page)
Topics (in English)
Chair Hidekata Hontani(Nagoya Inst. of Tech.)
Vice Chair Hideaki Haneishi(Chiba Univ.) / Takayuki Kitasaka(Aichi Inst. of Tech.)
Secretary Hideaki Haneishi(Yamaguchi Univ.) / Takayuki Kitasaka(Univ. of Hyogo)
Assistant Takeshi Hara(Gifu Univ.) / Yoshito Otake(NAIST)

Paper Information
Registration To Technical Committee on Medical Imaging
Language JPN
Title (in Japanese) (See Japanese page)
Sub Title (in Japanese) (See Japanese page)
Title (in English) Learning of Squamous Cell Image Classification Model Using Preference Learning to Assist Cervical Cytology
Sub Title (in English)
Keyword(1) Cell image classification
Keyword(2) Learning to rank
Keyword(3) Deep learning
1st Author's Name Yuta Nambu
1st Author's Affiliation Graduate School of Media Architecture, Future University Hakodate, Hakodate, Japan.(Future Univ. Hakodate)
2nd Author's Name Tasuku Mariya
2nd Author's Affiliation Department of Obstetrics and Gynecology, Sapporo Medical University School of Medicine, Sapporo, Japan.(Sapporo Medical Univ.)
3rd Author's Name Syota Shinkai
3rd Author's Affiliation Department of Obstetrics and Gynecology, Sapporo Medical University School of Medicine, Sapporo, Japan.(Sapporo Medical Univ.)
4th Author's Name Mina Umemoto
4th Author's Affiliation Department of Obstetrics and Gynecology, Sapporo Medical University School of Medicine, Sapporo, Japan.(Sapporo Medical Univ.)
5th Author's Name Hiroko Asanuma
5th Author's Affiliation Department of Pathology 1st, Sapporo Medical University School of Medicine, Sapporo, Japan.(Sapporo Medical Univ.)
6th Author's Name Yoshihiko Hirohashi
6th Author's Affiliation Department of Pathology 1st, Sapporo Medical University School of Medicine, Sapporo, Japan.(Sapporo Medical Univ.)
7th Author's Name Tsuyoshi Saito
7th Author's Affiliation Department of Obstetrics and Gynecology, Sapporo Medical University School of Medicine, Sapporo, Japan.(Sapporo Medical Univ.)
8th Author's Name Toshihiko Torigoe
8th Author's Affiliation Department of Pathology 1st, Sapporo Medical University School of Medicine, Sapporo, Japan.(Sapporo Medical Univ.)
9th Author's Name Ikuma Sato
9th Author's Affiliation Department of Media Architecture, Future University Hakodate, Hakodate, Japan.(Future Univ. Hakodate)
10th Author's Name Yuichi Fujino
10th Author's Affiliation Department of Media Architecture, Future University Hakodate, Hakodate, Japan.(Future Univ. Hakodate)
Date 2022-09-15
Paper # MI2022-62
Volume (vol) vol.122
Number (no) MI-188
Page pp.pp.53-58(MI),
#Pages 6
Date of Issue 2022-09-08 (MI)