A noninvasive model for chronic kidney disease screening and common pathological type identification from retinal images

Qianni Wu; Jianbo Li; Dong Liu; Jingyi Wen; Yunuo Wang; Yiqin Wang; Naya Huang; Lanping Jiang; Qinghua Liu; Hanming Lin; Pengxia Wan; Shicong Yang; Wenfang Chen; Hongjian Ye; Mohammed Haji Rashid Hassan; Ahmed Hassan Nur; Zefang Dai; Jie Guo; Shanshan Zhou; Jianwen Yu; Weixing Zhang; Wenben Chen; Ruiyang Li; Wai Cheng Iao; Juan-juan Feng; Yan Wang; Hua Hong; Peihong Yin; Qing Ye; Chao Xie; Min Zhu; Xiaoyi Liu; Yaozhong Kong; Jie Wang; Ruiying Ma; Yu Xiao; Guoguang Chen; Rongguo Fu; Yuhe Ke; Jasmine Ong Chiat Ling; Charumathi Sabanayagam; Daniel Shu Wei Ting; Kar Keung Cheng; Duoru Lin; Wei Chen; Haotian Lin

doi:10.1038/s41467-025-62273-0

A noninvasive model for chronic kidney disease screening and common pathological type identification from retinal images

Qianni Wu
, Jianbo Li
, Dong Liu
, Jingyi Wen
, Yunuo Wang
, Yiqin Wang
, Naya Huang
, Lanping Jiang
, Qinghua Liu
, Hanming Lin
, Pengxia Wan
, Shicong Yang
, Wenfang Chen
, Hongjian Ye
, Mohammed Haji Rashid Hassan
, Ahmed Hassan Nur
, Zefang Dai
, Jie Guo
, Shanshan Zhou
, Jianwen Yu

Weixing Zhang, Wenben Chen, Ruiyang Li, Wai Cheng Iao, Juan-juan Feng, Yan Wang, Hua Hong, Peihong Yin, Qing Ye, Chao Xie, Min Zhu, Xiaoyi Liu, Yaozhong Kong, Jie Wang, Ruiying Ma, Yu Xiao, Guoguang Chen, Rongguo Fu, Yuhe Ke, Jasmine Ong Chiat Ling, Charumathi Sabanayagam, Daniel Shu Wei Ting, Kar Keung Cheng, Duoru Lin^*, Wei Chen^*, Haotian Lin^*

^*Corresponding author for this work

Applied Health Sciences

Research output: Contribution to journal › Article › peer-review

69 Downloads (Pure)

Abstract

Chronic kidney disease (CKD) is a global health challenge, but invasive renal biopsies, the gold standard for diagnosis and prognosis, are often clinically constrained. To address this, we developed the kidney intelligent diagnosis system (KIDS), a noninvasive model for renal biopsy prediction using 13,144 retinal images from 6773 participants. The KIDS achieves an area under the receiver operating characteristic curve (AUC) of 0.839–0.993 for CKD screening and accurately identifies the five most common pathological types (AUC: 0.790–0.932) in a multicenter and multi-ethnic validation, outperforming nephrologists by 26.98% in accuracy. Additionally, the KIDS further predicts disease progression based on pathological classification. Given its flexible strategy, the KIDS can be adapted to local conditions to provide a tailored tool for patients. This noninvasive model has the potential to improve CKD clinical management, particularly for those who are ineligible for biopsies.

Original language	English
Article number	6962
Number of pages	15
Journal	Nature Communications
Volume	16
DOIs	https://doi.org/10.1038/s41467-025-62273-0
Publication status	Published - 29 Jul 2025

Access to Document

10.1038/s41467-025-62273-0Licence: Creative Commons: Attribution-NonCommercial-NoDerivs (CC BY-NC-ND)

WuQ2025noninvaisiveFinal published version, 2.49 MBLicence: Creative Commons: Attribution-NonCommercial-NoDerivs (CC BY-NC-ND)

Author Correction: A noninvasive model for chronic kidney disease screening and common pathological type identification from retinal images
Wu, Q., Li, J., Liu, D., Wen, J., Wang, Y., Wang, Y., Huang, N., Jiang, L., Liu, Q., Lin, H., Wan, P., Yang, S., Chen, W., Ye, H., Rashid Hassan, M. H., Nur, A. H., Dai, Z., Guo, J., Zhou, S. & Yu, J. & 26 others, Zhang, W., Chen, W., Li, R., Iao, W. C., Feng, J.-J., Wang, Y., Hong, H., Yin, P., Ye, Q., Xie, C., Zhu, M., Liu, X., Kong, Y., Wang, J., Ma, R., Xiao, Y., Chen, G., Fu, R., Ke, Y., Ling, J. O. C., Sabanayagam, C., Ting, D. S. W., Cheng, K. K., Lin, D., Chen, W. & Lin, H., 22 Sept 2025, In: Nature Communications. 16, 1 p., 8333.
Research output: Contribution to journal › Comment/debate

Open Access
File
8 Downloads (Pure)

Cite this

Wu, Q., Li, J., Liu, D., Wen, J., Wang, Y., Wang, Y., Huang, N., Jiang, L., Liu, Q., Lin, H., Wan, P., Yang, S., Chen, W., Ye, H., Rashid Hassan, M. H., Nur, A. H., Dai, Z., Guo, J., Zhou, S., ... Lin, H. (2025). A noninvasive model for chronic kidney disease screening and common pathological type identification from retinal images. Nature Communications, 16, Article 6962. https://doi.org/10.1038/s41467-025-62273-0

@article{ad6c1d015ea4425497deb64bc07c5321,

title = "A noninvasive model for chronic kidney disease screening and common pathological type identification from retinal images",

abstract = "Chronic kidney disease (CKD) is a global health challenge, but invasive renal biopsies, the gold standard for diagnosis and prognosis, are often clinically constrained. To address this, we developed the kidney intelligent diagnosis system (KIDS), a noninvasive model for renal biopsy prediction using 13,144 retinal images from 6773 participants. The KIDS achieves an area under the receiver operating characteristic curve (AUC) of 0.839–0.993 for CKD screening and accurately identifies the five most common pathological types (AUC: 0.790–0.932) in a multicenter and multi-ethnic validation, outperforming nephrologists by 26.98\% in accuracy. Additionally, the KIDS further predicts disease progression based on pathological classification. Given its flexible strategy, the KIDS can be adapted to local conditions to provide a tailored tool for patients. This noninvasive model has the potential to improve CKD clinical management, particularly for those who are ineligible for biopsies.",

author = "Qianni Wu and Jianbo Li and Dong Liu and Jingyi Wen and Yunuo Wang and Yiqin Wang and Naya Huang and Lanping Jiang and Qinghua Liu and Hanming Lin and Pengxia Wan and Shicong Yang and Wenfang Chen and Hongjian Ye and \{Rashid Hassan\}, \{Mohammed Haji\} and Nur, \{Ahmed Hassan\} and Zefang Dai and Jie Guo and Shanshan Zhou and Jianwen Yu and Weixing Zhang and Wenben Chen and Ruiyang Li and Iao, \{Wai Cheng\} and Juan-juan Feng and Yan Wang and Hua Hong and Peihong Yin and Qing Ye and Chao Xie and Min Zhu and Xiaoyi Liu and Yaozhong Kong and Jie Wang and Ruiying Ma and Yu Xiao and Guoguang Chen and Rongguo Fu and Yuhe Ke and Ling, \{Jasmine Ong Chiat\} and Charumathi Sabanayagam and Ting, \{Daniel Shu Wei\} and Cheng, \{Kar Keung\} and Duoru Lin and Wei Chen and Haotian Lin",

year = "2025",

month = jul,

day = "29",

doi = "10.1038/s41467-025-62273-0",

language = "English",

volume = "16",

journal = "Nature Communications",

issn = "2041-1723",

publisher = "Springer",

}

Wu, Q, Li, J, Liu, D, Wen, J, Wang, Y, Wang, Y, Huang, N, Jiang, L, Liu, Q, Lin, H, Wan, P, Yang, S, Chen, W, Ye, H, Rashid Hassan, MH, Nur, AH, Dai, Z, Guo, J, Zhou, S, Yu, J, Zhang, W, Chen, W, Li, R, Iao, WC, Feng, J, Wang, Y, Hong, H, Yin, P, Ye, Q, Xie, C, Zhu, M, Liu, X, Kong, Y, Wang, J, Ma, R, Xiao, Y, Chen, G, Fu, R, Ke, Y, Ling, JOC, Sabanayagam, C, Ting, DSW, Cheng, KK, Lin, D, Chen, W & Lin, H 2025, 'A noninvasive model for chronic kidney disease screening and common pathological type identification from retinal images', Nature Communications, vol. 16, 6962. https://doi.org/10.1038/s41467-025-62273-0

TY - JOUR

T1 - A noninvasive model for chronic kidney disease screening and common pathological type identification from retinal images

AU - Wu, Qianni

AU - Li, Jianbo

AU - Liu, Dong

AU - Wen, Jingyi

AU - Wang, Yunuo

AU - Wang, Yiqin

AU - Huang, Naya

AU - Jiang, Lanping

AU - Liu, Qinghua

AU - Lin, Hanming

AU - Wan, Pengxia

AU - Yang, Shicong

AU - Chen, Wenfang

AU - Ye, Hongjian

AU - Rashid Hassan, Mohammed Haji

AU - Nur, Ahmed Hassan

AU - Dai, Zefang

AU - Guo, Jie

AU - Zhou, Shanshan

AU - Yu, Jianwen

AU - Zhang, Weixing

AU - Chen, Wenben

AU - Li, Ruiyang

AU - Iao, Wai Cheng

AU - Feng, Juan-juan

AU - Wang, Yan

AU - Hong, Hua

AU - Yin, Peihong

AU - Ye, Qing

AU - Xie, Chao

AU - Zhu, Min

AU - Liu, Xiaoyi

AU - Kong, Yaozhong

AU - Wang, Jie

AU - Ma, Ruiying

AU - Xiao, Yu

AU - Chen, Guoguang

AU - Fu, Rongguo

AU - Ke, Yuhe

AU - Ling, Jasmine Ong Chiat

AU - Sabanayagam, Charumathi

AU - Ting, Daniel Shu Wei

AU - Cheng, Kar Keung

AU - Lin, Duoru

AU - Chen, Wei

AU - Lin, Haotian

PY - 2025/7/29

Y1 - 2025/7/29

N2 - Chronic kidney disease (CKD) is a global health challenge, but invasive renal biopsies, the gold standard for diagnosis and prognosis, are often clinically constrained. To address this, we developed the kidney intelligent diagnosis system (KIDS), a noninvasive model for renal biopsy prediction using 13,144 retinal images from 6773 participants. The KIDS achieves an area under the receiver operating characteristic curve (AUC) of 0.839–0.993 for CKD screening and accurately identifies the five most common pathological types (AUC: 0.790–0.932) in a multicenter and multi-ethnic validation, outperforming nephrologists by 26.98% in accuracy. Additionally, the KIDS further predicts disease progression based on pathological classification. Given its flexible strategy, the KIDS can be adapted to local conditions to provide a tailored tool for patients. This noninvasive model has the potential to improve CKD clinical management, particularly for those who are ineligible for biopsies.

AB - Chronic kidney disease (CKD) is a global health challenge, but invasive renal biopsies, the gold standard for diagnosis and prognosis, are often clinically constrained. To address this, we developed the kidney intelligent diagnosis system (KIDS), a noninvasive model for renal biopsy prediction using 13,144 retinal images from 6773 participants. The KIDS achieves an area under the receiver operating characteristic curve (AUC) of 0.839–0.993 for CKD screening and accurately identifies the five most common pathological types (AUC: 0.790–0.932) in a multicenter and multi-ethnic validation, outperforming nephrologists by 26.98% in accuracy. Additionally, the KIDS further predicts disease progression based on pathological classification. Given its flexible strategy, the KIDS can be adapted to local conditions to provide a tailored tool for patients. This noninvasive model has the potential to improve CKD clinical management, particularly for those who are ineligible for biopsies.

U2 - 10.1038/s41467-025-62273-0

DO - 10.1038/s41467-025-62273-0

M3 - Article

SN - 2041-1723

VL - 16

JO - Nature Communications

JF - Nature Communications

M1 - 6962

ER -

A noninvasive model for chronic kidney disease screening and common pathological type identification from retinal images

Abstract

Access to Document

Fingerprint

Research output

Author Correction: A noninvasive model for chronic kidney disease screening and common pathological type identification from retinal images

Cite this