甲襞毛细血管镜检测与深度学习在糖尿病中的应用
Abstract Objective To determine whether nailfold capillary images, acquired using video capillaroscopy, can provide diagnostic information about diabetes and its complications. Research Design and Methods Nailfold video capillaroscopy was performed in 120 adult patients with and without type 1 or ty...
| Main Authors: | , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Wiley
2023-02-01
|
| Series: | Journal of Diabetes |
| Subjects: | |
| Online Access: | https://doi.org/10.1111/1753-0407.13354 |
| _version_ | 1828013880766365696 |
|---|---|
| author | Reema Shah Jeremy Petch Walter Nelson Karsten Roth Michael D. Noseworthy Marzyeh Ghassemi Hertzel C. Gerstein |
| author_facet | Reema Shah Jeremy Petch Walter Nelson Karsten Roth Michael D. Noseworthy Marzyeh Ghassemi Hertzel C. Gerstein |
| author_sort | Reema Shah |
| collection | DOAJ |
| description | Abstract Objective To determine whether nailfold capillary images, acquired using video capillaroscopy, can provide diagnostic information about diabetes and its complications. Research Design and Methods Nailfold video capillaroscopy was performed in 120 adult patients with and without type 1 or type 2 diabetes, and with and without cardiovascular disease. Nailfold images were analyzed using convolutional neural networks, a deep learning technique. Cross‐validation was used to develop and test the ability of models to predict five5 prespecified states (diabetes, high glycosylated hemoglobin, cardiovascular event, retinopathy, albuminuria, and hypertension). The performance of each model for a particular state was assessed by estimating areas under the receiver operating characteristics curves (AUROC) and precision recall curves (AUPR). Results A total of 5236 nailfold images were acquired from 120 participants (mean 44 images per participant) and were all available for analysis. Models were able to accurately identify the presence of diabetes, with AUROC 0.84 (95% confidence interval [CI] 0.76, 0.91) and AUPR 0.84 (95% CI 0.78, 0.93), respectively. Models were also able to predict a history of cardiovascular events in patients with diabetes, with AUROC 0.65 (95% CI 0.51, 0.78) and AUPR 0.72 (95% CI 0.62, 0.88) respectively. Conclusions This proof‐of‐concept study demonstrates the potential of machine learning for identifying people with microvascular capillary changes from diabetes based on nailfold images, and for possibly identifying those most likely to have diabetes‐related complications. |
| first_indexed | 2024-04-10T09:51:51Z |
| format | Article |
| id | doaj.art-3edc2cf856fb4e98af7fc88e87d80df6 |
| institution | Directory Open Access Journal |
| issn | 1753-0393 1753-0407 |
| language | English |
| last_indexed | 2024-04-10T09:51:51Z |
| publishDate | 2023-02-01 |
| publisher | Wiley |
| record_format | Article |
| series | Journal of Diabetes |
| spelling | doaj.art-3edc2cf856fb4e98af7fc88e87d80df62023-02-16T23:09:18ZengWileyJournal of Diabetes1753-03931753-04072023-02-0115214515110.1111/1753-0407.13354甲襞毛细血管镜检测与深度学习在糖尿病中的应用Reema Shah0Jeremy Petch1Walter Nelson2Karsten Roth3Michael D. Noseworthy4Marzyeh Ghassemi5Hertzel C. Gerstein6Population Health Research Institute, McMaster University and Hamilton Health Sciences Hamilton Ontario CanadaPopulation Health Research Institute, McMaster University and Hamilton Health Sciences Hamilton Ontario CanadaCentre for Data Science and Digital Health Hamilton Health Sciences Hamilton Ontario CanadaCluster of Excellence Machine Learning University of Tübingen Tübingen GermanyElectrical and Computer Engineering McMaster University Hamilton Ontario CanadaVector Institute Toronto Ontario CanadaPopulation Health Research Institute, McMaster University and Hamilton Health Sciences Hamilton Ontario CanadaAbstract Objective To determine whether nailfold capillary images, acquired using video capillaroscopy, can provide diagnostic information about diabetes and its complications. Research Design and Methods Nailfold video capillaroscopy was performed in 120 adult patients with and without type 1 or type 2 diabetes, and with and without cardiovascular disease. Nailfold images were analyzed using convolutional neural networks, a deep learning technique. Cross‐validation was used to develop and test the ability of models to predict five5 prespecified states (diabetes, high glycosylated hemoglobin, cardiovascular event, retinopathy, albuminuria, and hypertension). The performance of each model for a particular state was assessed by estimating areas under the receiver operating characteristics curves (AUROC) and precision recall curves (AUPR). Results A total of 5236 nailfold images were acquired from 120 participants (mean 44 images per participant) and were all available for analysis. Models were able to accurately identify the presence of diabetes, with AUROC 0.84 (95% confidence interval [CI] 0.76, 0.91) and AUPR 0.84 (95% CI 0.78, 0.93), respectively. Models were also able to predict a history of cardiovascular events in patients with diabetes, with AUROC 0.65 (95% CI 0.51, 0.78) and AUPR 0.72 (95% CI 0.62, 0.88) respectively. Conclusions This proof‐of‐concept study demonstrates the potential of machine learning for identifying people with microvascular capillary changes from diabetes based on nailfold images, and for possibly identifying those most likely to have diabetes‐related complications.https://doi.org/10.1111/1753-0407.13354糖尿病诊断视网膜病变甲襞毛细血管机器学习 |
| spellingShingle | Reema Shah Jeremy Petch Walter Nelson Karsten Roth Michael D. Noseworthy Marzyeh Ghassemi Hertzel C. Gerstein 甲襞毛细血管镜检测与深度学习在糖尿病中的应用 Journal of Diabetes 糖尿病诊断 视网膜病变 甲襞毛细血管 机器学习 |
| title | 甲襞毛细血管镜检测与深度学习在糖尿病中的应用 |
| title_full | 甲襞毛细血管镜检测与深度学习在糖尿病中的应用 |
| title_fullStr | 甲襞毛细血管镜检测与深度学习在糖尿病中的应用 |
| title_full_unstemmed | 甲襞毛细血管镜检测与深度学习在糖尿病中的应用 |
| title_short | 甲襞毛细血管镜检测与深度学习在糖尿病中的应用 |
| title_sort | 甲襞毛细血管镜检测与深度学习在糖尿病中的应用 |
| topic | 糖尿病诊断 视网膜病变 甲襞毛细血管 机器学习 |
| url | https://doi.org/10.1111/1753-0407.13354 |
| work_keys_str_mv | AT reemashah jiǎbìmáoxìxuèguǎnjìngjiǎncèyǔshēndùxuéxízàitángniàobìngzhōngdeyīngyòng AT jeremypetch jiǎbìmáoxìxuèguǎnjìngjiǎncèyǔshēndùxuéxízàitángniàobìngzhōngdeyīngyòng AT walternelson jiǎbìmáoxìxuèguǎnjìngjiǎncèyǔshēndùxuéxízàitángniàobìngzhōngdeyīngyòng AT karstenroth jiǎbìmáoxìxuèguǎnjìngjiǎncèyǔshēndùxuéxízàitángniàobìngzhōngdeyīngyòng AT michaeldnoseworthy jiǎbìmáoxìxuèguǎnjìngjiǎncèyǔshēndùxuéxízàitángniàobìngzhōngdeyīngyòng AT marzyehghassemi jiǎbìmáoxìxuèguǎnjìngjiǎncèyǔshēndùxuéxízàitángniàobìngzhōngdeyīngyòng AT hertzelcgerstein jiǎbìmáoxìxuèguǎnjìngjiǎncèyǔshēndùxuéxízàitángniàobìngzhōngdeyīngyòng |