Multimodal deep learning improving the accuracy of pathological diagnoses for membranous nephropathy

Objectives Renal biopsy is the gold standard for the diagnosis of glomerular diseases including membranous nephropathy (MN), however, it faces challenges in accuracy, objectivity, and reproducibility of tissue evaluation. This study aims to develop a multimodal pathological diagnosis system to assis...

Full description

Saved in:
Bibliographic Details
Main Authors: Xiuxiu Hu, Jinyue Yang, Yiping Li, Yuxiang Gong, Haifeng Ni, Qing Wei, Minyu Yang, Yu Zhang, Jing Huang, Cao Ma, Bizhen Wei, Kaijie Yu, Jiayun Xu, Siyu Xia, Taotao Tang, Pingsheng Chen
Format: Article
Language:English
Published: Taylor & Francis Group 2025-12-01
Series:Renal Failure
Subjects:
Online Access:https://www.tandfonline.com/doi/10.1080/0886022X.2025.2528106
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1850112502676848640
author Xiuxiu Hu
Jinyue Yang
Yiping Li
Yuxiang Gong
Haifeng Ni
Qing Wei
Minyu Yang
Yu Zhang
Jing Huang
Cao Ma
Bizhen Wei
Kaijie Yu
Jiayun Xu
Siyu Xia
Taotao Tang
Pingsheng Chen
author_facet Xiuxiu Hu
Jinyue Yang
Yiping Li
Yuxiang Gong
Haifeng Ni
Qing Wei
Minyu Yang
Yu Zhang
Jing Huang
Cao Ma
Bizhen Wei
Kaijie Yu
Jiayun Xu
Siyu Xia
Taotao Tang
Pingsheng Chen
author_sort Xiuxiu Hu
collection DOAJ
description Objectives Renal biopsy is the gold standard for the diagnosis of glomerular diseases including membranous nephropathy (MN), however, it faces challenges in accuracy, objectivity, and reproducibility of tissue evaluation. This study aims to develop a multimodal pathological diagnosis system to assist pathologists in diagnosing MN in morphology.Methods Using PASM-stained, immunofluorescence, and electron microscopy images from MN patients, we built three deep-learning models to detect lesions. The outputs of these models were combined to provide a comprehensive pathological diagnosis. Our system was compared with pathologists, validated on external test sets, and detected in 138 patients with various kidney diseases.Results Considering PASM-stained images, our model had a classification accuracy of 91.74%, a recall of 81.97%, and an F1 score of 86.58% for spike identification. For immunofluorescence images, our model had an accuracy rate of 98.97%, a recall rate of 99.65%, and an F1 score of 99.31% for MN classification. Regarding the segmentation of electron-dense deposits, the segmentation model had a Dice coefficient of 85.66% and an IoU of 75.93%. Our model presented superior performance to that of pathologists in fluorescence image classification and segmentation of deposits, achieved high accuracy in spike identification and fluorescence image classification in external test sets, and could be targeted to diagnose MN in a wide range of glomerular diseases.Conclusions This multimodal pathological diagnosis system can not only assist pathologists in diagnosing MN rapidly and accurately but also lays the foundation to develop diagnostic models for other glomerular diseases.
format Article
id doaj-art-4303eaa8de974f72bf05328bbe5c5c6d
institution OA Journals
issn 0886-022X
1525-6049
language English
publishDate 2025-12-01
publisher Taylor & Francis Group
record_format Article
series Renal Failure
spelling doaj-art-4303eaa8de974f72bf05328bbe5c5c6d2025-08-20T02:37:21ZengTaylor & Francis GroupRenal Failure0886-022X1525-60492025-12-0147110.1080/0886022X.2025.2528106Multimodal deep learning improving the accuracy of pathological diagnoses for membranous nephropathyXiuxiu Hu0Jinyue Yang1Yiping Li2Yuxiang Gong3Haifeng Ni4Qing Wei5Minyu Yang6Yu Zhang7Jing Huang8Cao Ma9Bizhen Wei10Kaijie Yu11Jiayun Xu12Siyu Xia13Taotao Tang14Pingsheng Chen15Department of Pathology, School of Medicine, Southeast University, Nanjing, ChinaSchool of Automation, Southeast University, Nanjing, ChinaDepartment of Pathology, School of Medicine, Southeast University, Nanjing, ChinaDepartment of Nephrology, Zhongda Hospital, Southeast University, Nanjing, ChinaDepartment of Nephrology, Zhongda Hospital, Southeast University, Nanjing, ChinaDepartment of Nephrology, Zhongda Hospital, Southeast University, Nanjing, ChinaDepartment of Nephrology, Zhongda Hospital, Southeast University, Nanjing, ChinaDepartment of Pathology, School of Medicine, Southeast University, Nanjing, ChinaDepartment of Respiratory and Critical Care Medicine, Zhongda Hospital, Southeast University, Nanjing, ChinaDepartment of Pathology, School of Medicine, Southeast University, Nanjing, ChinaDepartment of Pathology, School of Medicine, Southeast University, Nanjing, ChinaDepartment of Pathology, School of Medicine, Southeast University, Nanjing, ChinaDepartment of Nephrology, The First Affiliated Hospital of Henan University of Science and Technology, Luoyang, ChinaSchool of Automation, Southeast University, Nanjing, ChinaDepartment of Pathology, School of Medicine, Southeast University, Nanjing, ChinaDepartment of Pathology, School of Medicine, Southeast University, Nanjing, ChinaObjectives Renal biopsy is the gold standard for the diagnosis of glomerular diseases including membranous nephropathy (MN), however, it faces challenges in accuracy, objectivity, and reproducibility of tissue evaluation. This study aims to develop a multimodal pathological diagnosis system to assist pathologists in diagnosing MN in morphology.Methods Using PASM-stained, immunofluorescence, and electron microscopy images from MN patients, we built three deep-learning models to detect lesions. The outputs of these models were combined to provide a comprehensive pathological diagnosis. Our system was compared with pathologists, validated on external test sets, and detected in 138 patients with various kidney diseases.Results Considering PASM-stained images, our model had a classification accuracy of 91.74%, a recall of 81.97%, and an F1 score of 86.58% for spike identification. For immunofluorescence images, our model had an accuracy rate of 98.97%, a recall rate of 99.65%, and an F1 score of 99.31% for MN classification. Regarding the segmentation of electron-dense deposits, the segmentation model had a Dice coefficient of 85.66% and an IoU of 75.93%. Our model presented superior performance to that of pathologists in fluorescence image classification and segmentation of deposits, achieved high accuracy in spike identification and fluorescence image classification in external test sets, and could be targeted to diagnose MN in a wide range of glomerular diseases.Conclusions This multimodal pathological diagnosis system can not only assist pathologists in diagnosing MN rapidly and accurately but also lays the foundation to develop diagnostic models for other glomerular diseases.https://www.tandfonline.com/doi/10.1080/0886022X.2025.2528106Membranous nephropathyrenal biopsymultimodal pathological diagnosisartificial intelligencedeep learning
spellingShingle Xiuxiu Hu
Jinyue Yang
Yiping Li
Yuxiang Gong
Haifeng Ni
Qing Wei
Minyu Yang
Yu Zhang
Jing Huang
Cao Ma
Bizhen Wei
Kaijie Yu
Jiayun Xu
Siyu Xia
Taotao Tang
Pingsheng Chen
Multimodal deep learning improving the accuracy of pathological diagnoses for membranous nephropathy
Renal Failure
Membranous nephropathy
renal biopsy
multimodal pathological diagnosis
artificial intelligence
deep learning
title Multimodal deep learning improving the accuracy of pathological diagnoses for membranous nephropathy
title_full Multimodal deep learning improving the accuracy of pathological diagnoses for membranous nephropathy
title_fullStr Multimodal deep learning improving the accuracy of pathological diagnoses for membranous nephropathy
title_full_unstemmed Multimodal deep learning improving the accuracy of pathological diagnoses for membranous nephropathy
title_short Multimodal deep learning improving the accuracy of pathological diagnoses for membranous nephropathy
title_sort multimodal deep learning improving the accuracy of pathological diagnoses for membranous nephropathy
topic Membranous nephropathy
renal biopsy
multimodal pathological diagnosis
artificial intelligence
deep learning
url https://www.tandfonline.com/doi/10.1080/0886022X.2025.2528106
work_keys_str_mv AT xiuxiuhu multimodaldeeplearningimprovingtheaccuracyofpathologicaldiagnosesformembranousnephropathy
AT jinyueyang multimodaldeeplearningimprovingtheaccuracyofpathologicaldiagnosesformembranousnephropathy
AT yipingli multimodaldeeplearningimprovingtheaccuracyofpathologicaldiagnosesformembranousnephropathy
AT yuxianggong multimodaldeeplearningimprovingtheaccuracyofpathologicaldiagnosesformembranousnephropathy
AT haifengni multimodaldeeplearningimprovingtheaccuracyofpathologicaldiagnosesformembranousnephropathy
AT qingwei multimodaldeeplearningimprovingtheaccuracyofpathologicaldiagnosesformembranousnephropathy
AT minyuyang multimodaldeeplearningimprovingtheaccuracyofpathologicaldiagnosesformembranousnephropathy
AT yuzhang multimodaldeeplearningimprovingtheaccuracyofpathologicaldiagnosesformembranousnephropathy
AT jinghuang multimodaldeeplearningimprovingtheaccuracyofpathologicaldiagnosesformembranousnephropathy
AT caoma multimodaldeeplearningimprovingtheaccuracyofpathologicaldiagnosesformembranousnephropathy
AT bizhenwei multimodaldeeplearningimprovingtheaccuracyofpathologicaldiagnosesformembranousnephropathy
AT kaijieyu multimodaldeeplearningimprovingtheaccuracyofpathologicaldiagnosesformembranousnephropathy
AT jiayunxu multimodaldeeplearningimprovingtheaccuracyofpathologicaldiagnosesformembranousnephropathy
AT siyuxia multimodaldeeplearningimprovingtheaccuracyofpathologicaldiagnosesformembranousnephropathy
AT taotaotang multimodaldeeplearningimprovingtheaccuracyofpathologicaldiagnosesformembranousnephropathy
AT pingshengchen multimodaldeeplearningimprovingtheaccuracyofpathologicaldiagnosesformembranousnephropathy