Authors

K. S. Wang, Central South University
G. Yu, School of Basic Medical Science Central South University
C. Xu, University of Oklahoma Health Sciences Center
X. H. Meng, Hunan Normal University
J. Zhou, Central South University
C. Zheng, Central South University
Z. Deng, Central South University
L. Shang, Central South University
R. Liu, Central South University
S. Su, Central South University
X. Zhou, Central South University
Q. Li, Central South University
J. Li, Central South University
J. Wang, Central South University
K. Ma, School of Basic Medical Science Central South University
J. Qi, School of Basic Medical Science Central South University
Z. Hu, School of Basic Medical Science Central South University
P. Tang, School of Basic Medical Science Central South University
J. Deng, Tulane University School of Medicine
X. Qiu, School of Basic Medical Science Central South University
B. Y. Li, School of Basic Medical Science Central South University
W. D. Shen, School of Basic Medical Science Central South University
R. P. Quan, School of Basic Medical Science Central South University
J. T. Yang, School of Basic Medical Science Central South University
L. Y. Huang, School of Basic Medical Science Central South University
Y. Xiao, School of Basic Medical Science Central South University
Z. C. Yang, Central South University
Z. Li, Central South University
S. C. Wang, Hunan Normal University
H. Ren, Second Military Medical University
C. Liang, Pathological Laboratory of Adicon Medical Laboratory Co., Ltd
W. Guo, Hunan Normal University
Y. Li, First Affiliated Hospital of Hunan Normal University
H. Xiao, Central South University
Y. Gu, Central South University
J.P. Yun, Sun Yat-Sen University Cancer Center
D. Huang, Fudan University Shanghai Cancer
Z. Song, Chinese PLA General Hospital
X. Fan, Nanjing Drum Tower Hospital
L. Chen, Air Force Medical University
X. Yan, Third Military Medical University
Z. Li, Guangdong Academy of Medical Sciences
Z.C. Huang, Central South University
J. Huang, Central South University
J. Luttrell, University of Southern Mississippi
Chaoyang Zhang, University of Southern MississippiFollow
W. Zhou, Michigan Technological University
K. Zhang, Xavier University of Louisiana
C. Yi, Ochsner Medical Center
C. Wu, Florida State University
H. Shen, Tulane University School of Medicine
Y.P. Wang, Tulane University School of Medicine
H.M. Xiao, Central South University
H.W. Deng, Tulane University School of Medicine

Document Type

Article

Publication Date

12-1-2021

School

Computing Sciences and Computer Engineering

Abstract

Background: Accurate and robust pathological image analysis for colorectal cancer (CRC) diagnosis is time-consuming and knowledge-intensive, but is essential for CRC patients’ treatment. The current heavy workload of pathologists in clinics/hospitals may easily lead to unconscious misdiagnosis of CRC based on daily image analyses.

Methods: Based on a state-of-the-art transfer-learned deep convolutional neural network in artificial intelligence (AI), we proposed a novel patch aggregation strategy for clinic CRC diagnosis using weakly labeled pathological whole-slide image (WSI) patches. This approach was trained and validated using an unprecedented and enormously large number of 170,099 patches, > 14,680 WSIs, from > 9631 subjects that covered diverse and representative clinical cases from multi-independent-sources across China, the USA, and Germany.

Results: Our innovative AI tool consistently and nearly perfectly agreed with (average Kappa statistic 0.896) and even often better than most of the experienced expert pathologists when tested in diagnosing CRC WSIs from multicenters. The average area under the receiver operating characteristics curve (AUC) of AI was greater than that of the pathologists (0.988 vs 0.970) and achieved the best performance among the application of other AI methods to CRC diagnosis. Our AI-generated heatmap highlights the image regions of cancer tissue/cells.

Conclusions: This first-ever generalizable AI system can handle large amounts of WSIs consistently and robustly without potential bias due to fatigue commonly experienced by clinical pathologists. It will drastically alleviate the heavy clinical burden of daily pathology diagnosis and improve the treatment for CRC patients. This tool is generalizable to other cancer diagnosis based on image recognition.

Comments

© BMC Medicine. This article is licensed under a Creative Commons Attribution 4.0 International License. Published version found at 10.1186/s12916-021-01942-5.

Publication Title

BMC Medicine

Volume

19

Issue

1

Find in your library

Share

COinS