Automated acquisition of explainable knowledge from unannotated histopathology images

Yoichiro Yamamoto (), Toyonori Tsuzuki, Jun Akatsuka, Masao Ueki, Hiromu Morikawa, Yasushi Numata, Taishi Takahara, Takuji Tsuyuki, Kotaro Tsutsumi, Ryuto Nakazawa, Akira Shimizu, Ichiro Maeda, Shinichi Tsuchiya, Hiroyuki Kanno, Yukihiro Kondo, Manabu Fukumoto, Gen Tamiya, Naonori Ueda and Go Kimura ()
Additional contact information
Yoichiro Yamamoto: RIKEN Center for Advanced Intelligence Project
Toyonori Tsuzuki: Aichi Medical University Hospital
Jun Akatsuka: RIKEN Center for Advanced Intelligence Project
Masao Ueki: RIKEN Center for Advanced Intelligence Project
Hiromu Morikawa: RIKEN Center for Advanced Intelligence Project
Yasushi Numata: RIKEN Center for Advanced Intelligence Project
Taishi Takahara: Aichi Medical University Hospital
Takuji Tsuyuki: Aichi Medical University Hospital
Kotaro Tsutsumi: RIKEN Center for Advanced Intelligence Project
Ryuto Nakazawa: St. Marianna University School of Medicine
Akira Shimizu: Nippon Medical School
Ichiro Maeda: RIKEN Center for Advanced Intelligence Project
Shinichi Tsuchiya: Ritsuzankai Iida Hospital
Hiroyuki Kanno: Shinshu University School of Medicine
Yukihiro Kondo: Nippon Medical School Hospital
Manabu Fukumoto: RIKEN Center for Advanced Intelligence Project
Gen Tamiya: RIKEN Center for Advanced Intelligence Project
Naonori Ueda: RIKEN Center for Advanced Intelligence Project
Go Kimura: Nippon Medical School Hospital

Nature Communications, 2019, vol. 10, issue 1, 1-9

Abstract: Abstract Deep learning algorithms have been successfully used in medical image classification. In the next stage, the technology of acquiring explainable knowledge from medical images is highly desired. Here we show that deep learning algorithm enables automated acquisition of explainable features from diagnostic annotation-free histopathology images. We compare the prediction accuracy of prostate cancer recurrence using our algorithm-generated features with that of diagnosis by expert pathologists using established criteria on 13,188 whole-mount pathology images consisting of over 86 billion image patches. Our method not only reveals findings established by humans but also features that have not been recognized, showing higher accuracy than human in prognostic prediction. Combining both our algorithm-generated features and human-established criteria predicts the recurrence more accurately than using either method alone. We confirm robustness of our method using external validation datasets including 2276 pathology images. This study opens up fields of machine learning analysis for discovering uncharted knowledge.

Date: 2019
References: Add references at CitEc
Citations:

Downloads: (external link)
https://www.nature.com/articles/s41467-019-13647-8 Abstract (text/html)

Related works:
This item may be available elsewhere in EconPapers: Search for items with the same title.

Export reference: BibTeX RIS (EndNote, ProCite, RefMan) HTML/Text

Persistent link: https://EconPapers.repec.org/RePEc:nat:natcom:v:10:y:2019:i:1:d:10.1038_s41467-019-13647-8

Ordering information: This journal article can be ordered from
https://www.nature.com/ncomms/

DOI: 10.1038/s41467-019-13647-8

Access Statistics for this article

Nature Communications is currently edited by Nathalie Le Bot, Enda Bergin and Fiona Gillespie

More articles in Nature Communications from Nature
Bibliographic data for series maintained by Sonal Shukla () and Springer Nature Abstracting and Indexing ().