Focusing on Clinically Interpretable Features: Selective Attention Regularization for Liver Biopsy Image Classification

Chong Yin; Siqi Liu; Rui Shao; Pong Chi Yuen

doi:10.1007/978-3-030-87240-3_15

Focusing on Clinically Interpretable Features: Selective Attention Regularization for Liver Biopsy Image Classification

Chong Yin, Siqi Liu, Rui Shao, Pong Chi Yuen^*

^*Corresponding author for this work

Department of Computer Science

Research output: Chapter in book/report/conference proceeding › Conference proceeding › peer-review

10 Citations (Scopus)

Abstract

Liver biopsy image analysis is the gold standard for early diagnosis of non-alcoholic fatty liver disease (NAFLD) worldwide. Deep neural networks offer an effective tool for image analysis. However, when applying deep learning methods to smaller histological image datasets, the model may be distracted by dominant normal tissues and ignore critical tissue alterations that pathologists focus on. In this paper, we propose a selective attention regularization module (SAttenReg) to mimic the diagnosis process of pathologists. Specifically, to explicitly encourage the model to focus on clinically interpretable features (e.g., nuclei and fat droplets), SAttenReg learns the attention map with the regularization of clinically interpretable features. Furthermore, with the different contributions of histological features, the model can selectively focus on different histological features based on the distribution of nuclei in each instance. Experiments conducted on the in-house Liver-NAS and public Biopsy4Grading biopsy image datasets show that our method achieves superior classification performance with promising localization results.

Original language	English
Title of host publication	Medical Image Computing and Computer Assisted Intervention – MICCAI 2021
Subtitle of host publication	24th International Conference, Strasbourg, France, September 27 – October 1, 2021, Proceedings, Part V
Editors	Marleen de Bruijne, Philippe C. Cattin, Stéphane Cotin, Nicolas Padoy, Stefanie Speidel, Yefeng Zheng, Caroline Essert
Publisher	Springer Cham
Pages	153-162
Number of pages	10
Edition	1st
ISBN (Electronic)	9783030872403
ISBN (Print)	9783030872397
DOIs	https://doi.org/10.1007/978-3-030-87240-3_15
Publication status	Published - 21 Sept 2021
Event	24th International Conference on Medical Image Computing and Computer-Assisted Intervention, MICCAI 2021 - Strasbourg, France Duration: 27 Sept 2021 → 1 Oct 2021

Publication series

Name	Lecture Notes in Computer Science
Volume	12905
ISSN (Print)	0302-9743
ISSN (Electronic)	1611-3349

Name	Image Processing, Computer Vision, Pattern Recognition, and Graphics
Volume	12905

Conference

Conference	24th International Conference on Medical Image Computing and Computer-Assisted Intervention, MICCAI 2021
Country/Territory	France
City	Strasbourg
Period	27/09/21 → 1/10/21

User-Defined Keywords

Liver biopsy images
Selective attention regularization

Access to Document

10.1007/978-3-030-87240-3_15

Cite this

Yin, C., Liu, S., Shao, R., & Yuen, P. C. (2021). Focusing on Clinically Interpretable Features: Selective Attention Regularization for Liver Biopsy Image Classification. In M. de Bruijne, P. C. Cattin, S. Cotin, N. Padoy, S. Speidel, Y. Zheng, & C. Essert (Eds.), Medical Image Computing and Computer Assisted Intervention – MICCAI 2021: 24th International Conference, Strasbourg, France, September 27 – October 1, 2021, Proceedings, Part V (1st ed., pp. 153-162). (Lecture Notes in Computer Science; Vol. 12905), (Image Processing, Computer Vision, Pattern Recognition, and Graphics; Vol. 12905). Springer Cham. https://doi.org/10.1007/978-3-030-87240-3_15

Yin, Chong ; Liu, Siqi ; Shao, Rui et al. / Focusing on Clinically Interpretable Features: Selective Attention Regularization for Liver Biopsy Image Classification. Medical Image Computing and Computer Assisted Intervention – MICCAI 2021: 24th International Conference, Strasbourg, France, September 27 – October 1, 2021, Proceedings, Part V. editor / Marleen de Bruijne ; Philippe C. Cattin ; Stéphane Cotin ; Nicolas Padoy ; Stefanie Speidel ; Yefeng Zheng ; Caroline Essert. 1st. ed. Springer Cham, 2021. pp. 153-162 (Lecture Notes in Computer Science). (Image Processing, Computer Vision, Pattern Recognition, and Graphics).

@inproceedings{d01b26afec8644eda87b2958a26b5def,

title = "Focusing on Clinically Interpretable Features: Selective Attention Regularization for Liver Biopsy Image Classification",

abstract = "Liver biopsy image analysis is the gold standard for early diagnosis of non-alcoholic fatty liver disease (NAFLD) worldwide. Deep neural networks offer an effective tool for image analysis. However, when applying deep learning methods to smaller histological image datasets, the model may be distracted by dominant normal tissues and ignore critical tissue alterations that pathologists focus on. In this paper, we propose a selective attention regularization module (SAttenReg) to mimic the diagnosis process of pathologists. Specifically, to explicitly encourage the model to focus on clinically interpretable features (e.g., nuclei and fat droplets), SAttenReg learns the attention map with the regularization of clinically interpretable features. Furthermore, with the different contributions of histological features, the model can selectively focus on different histological features based on the distribution of nuclei in each instance. Experiments conducted on the in-house Liver-NAS and public Biopsy4Grading biopsy image datasets show that our method achieves superior classification performance with promising localization results.",

keywords = "Liver biopsy images, Selective attention regularization",

author = "Chong Yin and Siqi Liu and Rui Shao and Yuen, {Pong Chi}",

note = "This work was supported by the Health and Medical Research Fund Project under Grant 07180216. {\textcopyright} 2021 Springer Nature Switzerland AG; 24th International Conference on Medical Image Computing and Computer-Assisted Intervention, MICCAI 2021 ; Conference date: 27-09-2021 Through 01-10-2021",

year = "2021",

month = sep,

day = "21",

doi = "10.1007/978-3-030-87240-3_15",

language = "English",

isbn = "9783030872397",

series = "Lecture Notes in Computer Science",

publisher = "Springer Cham",

pages = "153--162",

editor = "{de Bruijne}, Marleen and Cattin, {Philippe C.} and St{\'e}phane Cotin and Padoy, {Nicolas } and Stefanie Speidel and Yefeng Zheng and Caroline Essert",

booktitle = "Medical Image Computing and Computer Assisted Intervention – MICCAI 2021",

edition = "1st",

}

Yin, C, Liu, S, Shao, R & Yuen, PC 2021, Focusing on Clinically Interpretable Features: Selective Attention Regularization for Liver Biopsy Image Classification. in M de Bruijne, PC Cattin, S Cotin, N Padoy, S Speidel, Y Zheng & C Essert (eds), Medical Image Computing and Computer Assisted Intervention – MICCAI 2021: 24th International Conference, Strasbourg, France, September 27 – October 1, 2021, Proceedings, Part V. 1st edn, Lecture Notes in Computer Science, vol. 12905, Image Processing, Computer Vision, Pattern Recognition, and Graphics, vol. 12905, Springer Cham, pp. 153-162, 24th International Conference on Medical Image Computing and Computer-Assisted Intervention, MICCAI 2021, Strasbourg, France, 27/09/21. https://doi.org/10.1007/978-3-030-87240-3_15

Focusing on Clinically Interpretable Features: Selective Attention Regularization for Liver Biopsy Image Classification. / Yin, Chong; Liu, Siqi; Shao, Rui et al.
Medical Image Computing and Computer Assisted Intervention – MICCAI 2021: 24th International Conference, Strasbourg, France, September 27 – October 1, 2021, Proceedings, Part V. ed. / Marleen de Bruijne; Philippe C. Cattin; Stéphane Cotin; Nicolas Padoy; Stefanie Speidel; Yefeng Zheng; Caroline Essert. 1st. ed. Springer Cham, 2021. p. 153-162 (Lecture Notes in Computer Science; Vol. 12905), (Image Processing, Computer Vision, Pattern Recognition, and Graphics; Vol. 12905).

Research output: Chapter in book/report/conference proceeding › Conference proceeding › peer-review

TY - GEN

T1 - Focusing on Clinically Interpretable Features: Selective Attention Regularization for Liver Biopsy Image Classification

AU - Yin, Chong

AU - Liu, Siqi

AU - Shao, Rui

AU - Yuen, Pong Chi

PY - 2021/9/21

Y1 - 2021/9/21

N2 - Liver biopsy image analysis is the gold standard for early diagnosis of non-alcoholic fatty liver disease (NAFLD) worldwide. Deep neural networks offer an effective tool for image analysis. However, when applying deep learning methods to smaller histological image datasets, the model may be distracted by dominant normal tissues and ignore critical tissue alterations that pathologists focus on. In this paper, we propose a selective attention regularization module (SAttenReg) to mimic the diagnosis process of pathologists. Specifically, to explicitly encourage the model to focus on clinically interpretable features (e.g., nuclei and fat droplets), SAttenReg learns the attention map with the regularization of clinically interpretable features. Furthermore, with the different contributions of histological features, the model can selectively focus on different histological features based on the distribution of nuclei in each instance. Experiments conducted on the in-house Liver-NAS and public Biopsy4Grading biopsy image datasets show that our method achieves superior classification performance with promising localization results.

AB - Liver biopsy image analysis is the gold standard for early diagnosis of non-alcoholic fatty liver disease (NAFLD) worldwide. Deep neural networks offer an effective tool for image analysis. However, when applying deep learning methods to smaller histological image datasets, the model may be distracted by dominant normal tissues and ignore critical tissue alterations that pathologists focus on. In this paper, we propose a selective attention regularization module (SAttenReg) to mimic the diagnosis process of pathologists. Specifically, to explicitly encourage the model to focus on clinically interpretable features (e.g., nuclei and fat droplets), SAttenReg learns the attention map with the regularization of clinically interpretable features. Furthermore, with the different contributions of histological features, the model can selectively focus on different histological features based on the distribution of nuclei in each instance. Experiments conducted on the in-house Liver-NAS and public Biopsy4Grading biopsy image datasets show that our method achieves superior classification performance with promising localization results.

KW - Liver biopsy images

KW - Selective attention regularization

U2 - 10.1007/978-3-030-87240-3_15

DO - 10.1007/978-3-030-87240-3_15

M3 - Conference proceeding

SN - 9783030872397

T3 - Lecture Notes in Computer Science

SP - 153

EP - 162

BT - Medical Image Computing and Computer Assisted Intervention – MICCAI 2021

A2 - de Bruijne, Marleen

A2 - Cattin, Philippe C.

A2 - Cotin, Stéphane

A2 - Padoy, Nicolas

A2 - Speidel, Stefanie

A2 - Zheng, Yefeng

A2 - Essert, Caroline

PB - Springer Cham

T2 - 24th International Conference on Medical Image Computing and Computer-Assisted Intervention, MICCAI 2021

Y2 - 27 September 2021 through 1 October 2021

ER -

Yin C, Liu S, Shao R, Yuen PC. Focusing on Clinically Interpretable Features: Selective Attention Regularization for Liver Biopsy Image Classification. In de Bruijne M, Cattin PC, Cotin S, Padoy N, Speidel S, Zheng Y, Essert C, editors, Medical Image Computing and Computer Assisted Intervention – MICCAI 2021: 24th International Conference, Strasbourg, France, September 27 – October 1, 2021, Proceedings, Part V. 1st ed. Springer Cham. 2021. p. 153-162. (Lecture Notes in Computer Science). (Image Processing, Computer Vision, Pattern Recognition, and Graphics). doi: 10.1007/978-3-030-87240-3_15

Focusing on Clinically Interpretable Features: Selective Attention Regularization for Liver Biopsy Image Classification

Abstract

Publication series

Conference

User-Defined Keywords

Access to Document

Fingerprint

Cite this