DeepFundus : A flow-cytometry-like image quality classifier for boosting the whole life cycle of medical artificial intelligence
Copyright © 2022 The Authors. Published by Elsevier Inc. All rights reserved..
Medical artificial intelligence (AI) has been moving from the research phase to clinical implementation. However, most AI-based models are mainly built using high-quality images preprocessed in the laboratory, which is not representative of real-world settings. This dataset bias proves a major driver of AI system dysfunction. Inspired by the design of flow cytometry, DeepFundus, a deep-learning-based fundus image classifier, is developed to provide automated and multidimensional image sorting to address this data quality gap. DeepFundus achieves areas under the receiver operating characteristic curves (AUCs) over 0.9 in image classification concerning overall quality, clinical quality factors, and structural quality analysis on both the internal test and national validation datasets. Additionally, DeepFundus can be integrated into both model development and clinical application of AI diagnostics to significantly enhance model performance for detecting multiple retinopathies. DeepFundus can be used to construct a data-driven paradigm for improving the entire life cycle of medical AI practice.
Medienart: |
E-Artikel |
---|
Erscheinungsjahr: |
2023 |
---|---|
Erschienen: |
2023 |
Enthalten in: |
Zur Gesamtaufnahme - volume:4 |
---|---|
Enthalten in: |
Cell reports. Medicine - 4(2023), 2 vom: 21. Feb., Seite 100912 |
Sprache: |
Englisch |
---|
Beteiligte Personen: |
Liu, Lixue [VerfasserIn] |
---|
Links: |
---|
Themen: |
Artificial intelligence |
---|
Anmerkungen: |
Date Completed 24.02.2023 Date Revised 26.03.2023 published: Print-Electronic Citation Status MEDLINE |
---|
doi: |
10.1016/j.xcrm.2022.100912 |
---|
funding: |
|
---|---|
Förderinstitution / Projekttitel: |
|
PPN (Katalog-ID): |
NLM351802150 |
---|
LEADER | 01000naa a22002652 4500 | ||
---|---|---|---|
001 | NLM351802150 | ||
003 | DE-627 | ||
005 | 20231226052238.0 | ||
007 | cr uuu---uuuuu | ||
008 | 231226s2023 xx |||||o 00| ||eng c | ||
024 | 7 | |a 10.1016/j.xcrm.2022.100912 |2 doi | |
028 | 5 | 2 | |a pubmed24n1172.xml |
035 | |a (DE-627)NLM351802150 | ||
035 | |a (NLM)36669488 | ||
035 | |a (PII)S2666-3791(22)00491-8 | ||
040 | |a DE-627 |b ger |c DE-627 |e rakwb | ||
041 | |a eng | ||
100 | 1 | |a Liu, Lixue |e verfasserin |4 aut | |
245 | 1 | 0 | |a DeepFundus |b A flow-cytometry-like image quality classifier for boosting the whole life cycle of medical artificial intelligence |
264 | 1 | |c 2023 | |
336 | |a Text |b txt |2 rdacontent | ||
337 | |a ƒaComputermedien |b c |2 rdamedia | ||
338 | |a ƒa Online-Ressource |b cr |2 rdacarrier | ||
500 | |a Date Completed 24.02.2023 | ||
500 | |a Date Revised 26.03.2023 | ||
500 | |a published: Print-Electronic | ||
500 | |a Citation Status MEDLINE | ||
520 | |a Copyright © 2022 The Authors. Published by Elsevier Inc. All rights reserved. | ||
520 | |a Medical artificial intelligence (AI) has been moving from the research phase to clinical implementation. However, most AI-based models are mainly built using high-quality images preprocessed in the laboratory, which is not representative of real-world settings. This dataset bias proves a major driver of AI system dysfunction. Inspired by the design of flow cytometry, DeepFundus, a deep-learning-based fundus image classifier, is developed to provide automated and multidimensional image sorting to address this data quality gap. DeepFundus achieves areas under the receiver operating characteristic curves (AUCs) over 0.9 in image classification concerning overall quality, clinical quality factors, and structural quality analysis on both the internal test and national validation datasets. Additionally, DeepFundus can be integrated into both model development and clinical application of AI diagnostics to significantly enhance model performance for detecting multiple retinopathies. DeepFundus can be used to construct a data-driven paradigm for improving the entire life cycle of medical AI practice | ||
650 | 4 | |a Journal Article | |
650 | 4 | |a Research Support, Non-U.S. Gov't | |
650 | 4 | |a artificial intelligence | |
650 | 4 | |a image quality | |
650 | 4 | |a retinal diseases | |
700 | 1 | |a Wu, Xiaohang |e verfasserin |4 aut | |
700 | 1 | |a Lin, Duoru |e verfasserin |4 aut | |
700 | 1 | |a Zhao, Lanqin |e verfasserin |4 aut | |
700 | 1 | |a Li, Mingyuan |e verfasserin |4 aut | |
700 | 1 | |a Yun, Dongyuan |e verfasserin |4 aut | |
700 | 1 | |a Lin, Zhenzhe |e verfasserin |4 aut | |
700 | 1 | |a Pang, Jianyu |e verfasserin |4 aut | |
700 | 1 | |a Li, Longhui |e verfasserin |4 aut | |
700 | 1 | |a Wu, Yuxuan |e verfasserin |4 aut | |
700 | 1 | |a Lai, Weiyi |e verfasserin |4 aut | |
700 | 1 | |a Xiao, Wei |e verfasserin |4 aut | |
700 | 1 | |a Shang, Yuanjun |e verfasserin |4 aut | |
700 | 1 | |a Feng, Weibo |e verfasserin |4 aut | |
700 | 1 | |a Tan, Xiao |e verfasserin |4 aut | |
700 | 1 | |a Li, Qiang |e verfasserin |4 aut | |
700 | 1 | |a Liu, Shenzhen |e verfasserin |4 aut | |
700 | 1 | |a Lin, Xinxin |e verfasserin |4 aut | |
700 | 1 | |a Sun, Jiaxin |e verfasserin |4 aut | |
700 | 1 | |a Zhao, Yiqi |e verfasserin |4 aut | |
700 | 1 | |a Yang, Ximei |e verfasserin |4 aut | |
700 | 1 | |a Ye, Qinying |e verfasserin |4 aut | |
700 | 1 | |a Zhong, Yuesi |e verfasserin |4 aut | |
700 | 1 | |a Huang, Xi |e verfasserin |4 aut | |
700 | 1 | |a He, Yuan |e verfasserin |4 aut | |
700 | 1 | |a Fu, Ziwei |e verfasserin |4 aut | |
700 | 1 | |a Xiang, Yi |e verfasserin |4 aut | |
700 | 1 | |a Zhang, Li |e verfasserin |4 aut | |
700 | 1 | |a Zhao, Mingwei |e verfasserin |4 aut | |
700 | 1 | |a Qu, Jinfeng |e verfasserin |4 aut | |
700 | 1 | |a Xu, Fan |e verfasserin |4 aut | |
700 | 1 | |a Lu, Peng |e verfasserin |4 aut | |
700 | 1 | |a Li, Jianqiao |e verfasserin |4 aut | |
700 | 1 | |a Xu, Fabao |e verfasserin |4 aut | |
700 | 1 | |a Wei, Wenbin |e verfasserin |4 aut | |
700 | 1 | |a Dong, Li |e verfasserin |4 aut | |
700 | 1 | |a Dai, Guangzheng |e verfasserin |4 aut | |
700 | 1 | |a He, Xingru |e verfasserin |4 aut | |
700 | 1 | |a Yan, Wentao |e verfasserin |4 aut | |
700 | 1 | |a Zhu, Qiaolin |e verfasserin |4 aut | |
700 | 1 | |a Lu, Linna |e verfasserin |4 aut | |
700 | 1 | |a Zhang, Jiaying |e verfasserin |4 aut | |
700 | 1 | |a Zhou, Wei |e verfasserin |4 aut | |
700 | 1 | |a Meng, Xiangda |e verfasserin |4 aut | |
700 | 1 | |a Li, Shiying |e verfasserin |4 aut | |
700 | 1 | |a Shen, Mei |e verfasserin |4 aut | |
700 | 1 | |a Jiang, Qin |e verfasserin |4 aut | |
700 | 1 | |a Chen, Nan |e verfasserin |4 aut | |
700 | 1 | |a Zhou, Xingtao |e verfasserin |4 aut | |
700 | 1 | |a Li, Meiyan |e verfasserin |4 aut | |
700 | 1 | |a Wang, Yan |e verfasserin |4 aut | |
700 | 1 | |a Zou, Haohan |e verfasserin |4 aut | |
700 | 1 | |a Zhong, Hua |e verfasserin |4 aut | |
700 | 1 | |a Yang, Wenyan |e verfasserin |4 aut | |
700 | 1 | |a Shou, Wulin |e verfasserin |4 aut | |
700 | 1 | |a Zhong, Xingwu |e verfasserin |4 aut | |
700 | 1 | |a Yang, Zhenduo |e verfasserin |4 aut | |
700 | 1 | |a Ding, Lin |e verfasserin |4 aut | |
700 | 1 | |a Hu, Yongcheng |e verfasserin |4 aut | |
700 | 1 | |a Tan, Gang |e verfasserin |4 aut | |
700 | 1 | |a He, Wanji |e verfasserin |4 aut | |
700 | 1 | |a Zhao, Xin |e verfasserin |4 aut | |
700 | 1 | |a Chen, Yuzhong |e verfasserin |4 aut | |
700 | 1 | |a Liu, Yizhi |e verfasserin |4 aut | |
700 | 1 | |a Lin, Haotian |e verfasserin |4 aut | |
773 | 0 | 8 | |i Enthalten in |t Cell reports. Medicine |d 2020 |g 4(2023), 2 vom: 21. Feb., Seite 100912 |w (DE-627)NLM310587662 |x 2666-3791 |7 nnns |
773 | 1 | 8 | |g volume:4 |g year:2023 |g number:2 |g day:21 |g month:02 |g pages:100912 |
856 | 4 | 0 | |u http://dx.doi.org/10.1016/j.xcrm.2022.100912 |3 Volltext |
912 | |a GBV_USEFLAG_A | ||
912 | |a GBV_NLM | ||
951 | |a AR | ||
952 | |d 4 |j 2023 |e 2 |b 21 |c 02 |h 100912 |