Smartphone-Acquired Anterior Segment Images for Deep Learning Prediction of Anterior Chamber Depth: A Proof-of-Concept Study
PurposeTo develop a deep learning (DL) algorithm for predicting anterior chamber depth (ACD) from smartphone-acquired anterior segment photographs.MethodsFor algorithm development, we included 4,157 eyes from 2,084 Chinese primary school students (aged 11–15 years) from Mojiang Myopia Progression St...
| Main Authors: | , , , , , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Frontiers Media S.A.
2022-06-01
|
| Series: | Frontiers in Medicine |
| Subjects: | |
| Online Access: | https://www.frontiersin.org/articles/10.3389/fmed.2022.912214/full |
| _version_ | 1818251917891993600 |
|---|---|
| author | Chaoxu Qian Chaoxu Qian Yixing Jiang Zhi Da Soh Zhi Da Soh Ganesan Sakthi Selvam Shuyuan Xiao Yih-Chung Tham Yih-Chung Tham Yih-Chung Tham Xinxing Xu Yong Liu Jun Li Hua Zhong Ching-Yu Cheng Ching-Yu Cheng Ching-Yu Cheng |
| author_facet | Chaoxu Qian Chaoxu Qian Yixing Jiang Zhi Da Soh Zhi Da Soh Ganesan Sakthi Selvam Shuyuan Xiao Yih-Chung Tham Yih-Chung Tham Yih-Chung Tham Xinxing Xu Yong Liu Jun Li Hua Zhong Ching-Yu Cheng Ching-Yu Cheng Ching-Yu Cheng |
| author_sort | Chaoxu Qian |
| collection | DOAJ |
| description | PurposeTo develop a deep learning (DL) algorithm for predicting anterior chamber depth (ACD) from smartphone-acquired anterior segment photographs.MethodsFor algorithm development, we included 4,157 eyes from 2,084 Chinese primary school students (aged 11–15 years) from Mojiang Myopia Progression Study (MMPS). All participants had with ACD measurement measured with Lenstar (LS 900) and anterior segment photographs acquired from a smartphone (iPhone Xs), which was mounted on slit lamp and under diffuses lighting. The anterior segment photographs were randomly selected by person into training (80%, no. of eyes = 3,326) and testing (20%, no. of eyes = 831) dataset. We excluded participants with intraocular surgery history or pronounced corneal haze. A convolutional neural network was developed to predict ACD based on these anterior segment photographs. To determine the accuracy of our algorithm, we measured the mean absolute error (MAE) and coefficient of determination (R2) were evaluated. Bland Altman plot was used to illustrate the agreement between DL-predicted and measured ACD values.ResultsIn the test set of 831 eyes, the mean measured ACD was 3.06 ± 0.25 mm, and the mean DL-predicted ACD was 3.10 ± 0.20 mm. The MAE was 0.16 ± 0.13 mm, and R2 was 0.40 between the predicted and measured ACD. The overall mean difference was −0.04 ± 0.20 mm, with 95% limits of agreement ranging between −0.43 and 0.34 mm. The generated saliency maps showed that the algorithm mainly utilized central corneal region (i.e., the site where ACD is clinically measured typically) in making its prediction, providing further plausibility to the algorithm's prediction.ConclusionsWe developed a DL algorithm to estimate ACD based on smartphone-acquired anterior segment photographs. Upon further validation, our algorithm may be further refined for use as a ACD screening tool in rural localities where means of assessing ocular biometry is not readily available. This is particularly important in China where the risk of primary angle closure disease is high and often undetected. |
| first_indexed | 2024-12-12T16:15:54Z |
| format | Article |
| id | doaj.art-50cc51b41bbd450bb719bb93a5739f1d |
| institution | Directory Open Access Journal |
| issn | 2296-858X |
| language | English |
| last_indexed | 2024-12-12T16:15:54Z |
| publishDate | 2022-06-01 |
| publisher | Frontiers Media S.A. |
| record_format | Article |
| series | Frontiers in Medicine |
| spelling | doaj.art-50cc51b41bbd450bb719bb93a5739f1d2022-12-22T00:19:06ZengFrontiers Media S.A.Frontiers in Medicine2296-858X2022-06-01910.3389/fmed.2022.912214912214Smartphone-Acquired Anterior Segment Images for Deep Learning Prediction of Anterior Chamber Depth: A Proof-of-Concept StudyChaoxu Qian0Chaoxu Qian1Yixing Jiang2Zhi Da Soh3Zhi Da Soh4Ganesan Sakthi Selvam5Shuyuan Xiao6Yih-Chung Tham7Yih-Chung Tham8Yih-Chung Tham9Xinxing Xu10Yong Liu11Jun Li12Hua Zhong13Ching-Yu Cheng14Ching-Yu Cheng15Ching-Yu Cheng16Singapore Eye Research Institute, Singapore National Eye Centre, Singapore, SingaporeDepartment of Ophthalmology, The First Affiliated Hospital of Kunming Medical University, Kunming, ChinaInstitute of High Performance Computing, Agency for Science, Technology and Research (A*Star), Singapore, SingaporeSingapore Eye Research Institute, Singapore National Eye Centre, Singapore, SingaporeDepartment of Ophthalmology, Yong Loo Lin School of Medicine, National University of Singapore, Singapore, SingaporeInstitute of High Performance Computing, Agency for Science, Technology and Research (A*Star), Singapore, SingaporeDepartment of Ophthalmology, The First Affiliated Hospital of Kunming Medical University, Kunming, ChinaSingapore Eye Research Institute, Singapore National Eye Centre, Singapore, SingaporeDepartment of Ophthalmology, Yong Loo Lin School of Medicine, National University of Singapore, Singapore, SingaporeOphthalmology and Visual Sciences Academic Clinical Program (Eye ACP), Duke-NUS Medical School, Singapore, SingaporeInstitute of High Performance Computing, Agency for Science, Technology and Research (A*Star), Singapore, SingaporeInstitute of High Performance Computing, Agency for Science, Technology and Research (A*Star), Singapore, SingaporeDepartment of Ophthalmology, The Second People's Hospital of Yunnan Province, Kunming, ChinaDepartment of Ophthalmology, The First Affiliated Hospital of Kunming Medical University, Kunming, ChinaSingapore Eye Research Institute, Singapore National Eye Centre, Singapore, SingaporeDepartment of Ophthalmology, Yong Loo Lin School of Medicine, National University of Singapore, Singapore, SingaporeOphthalmology and Visual Sciences Academic Clinical Program (Eye ACP), Duke-NUS Medical School, Singapore, SingaporePurposeTo develop a deep learning (DL) algorithm for predicting anterior chamber depth (ACD) from smartphone-acquired anterior segment photographs.MethodsFor algorithm development, we included 4,157 eyes from 2,084 Chinese primary school students (aged 11–15 years) from Mojiang Myopia Progression Study (MMPS). All participants had with ACD measurement measured with Lenstar (LS 900) and anterior segment photographs acquired from a smartphone (iPhone Xs), which was mounted on slit lamp and under diffuses lighting. The anterior segment photographs were randomly selected by person into training (80%, no. of eyes = 3,326) and testing (20%, no. of eyes = 831) dataset. We excluded participants with intraocular surgery history or pronounced corneal haze. A convolutional neural network was developed to predict ACD based on these anterior segment photographs. To determine the accuracy of our algorithm, we measured the mean absolute error (MAE) and coefficient of determination (R2) were evaluated. Bland Altman plot was used to illustrate the agreement between DL-predicted and measured ACD values.ResultsIn the test set of 831 eyes, the mean measured ACD was 3.06 ± 0.25 mm, and the mean DL-predicted ACD was 3.10 ± 0.20 mm. The MAE was 0.16 ± 0.13 mm, and R2 was 0.40 between the predicted and measured ACD. The overall mean difference was −0.04 ± 0.20 mm, with 95% limits of agreement ranging between −0.43 and 0.34 mm. The generated saliency maps showed that the algorithm mainly utilized central corneal region (i.e., the site where ACD is clinically measured typically) in making its prediction, providing further plausibility to the algorithm's prediction.ConclusionsWe developed a DL algorithm to estimate ACD based on smartphone-acquired anterior segment photographs. Upon further validation, our algorithm may be further refined for use as a ACD screening tool in rural localities where means of assessing ocular biometry is not readily available. This is particularly important in China where the risk of primary angle closure disease is high and often undetected.https://www.frontiersin.org/articles/10.3389/fmed.2022.912214/fullprimary angle-closure glaucomaglaucomaanterior chamber depthsmartphonedeep learning |
| spellingShingle | Chaoxu Qian Chaoxu Qian Yixing Jiang Zhi Da Soh Zhi Da Soh Ganesan Sakthi Selvam Shuyuan Xiao Yih-Chung Tham Yih-Chung Tham Yih-Chung Tham Xinxing Xu Yong Liu Jun Li Hua Zhong Ching-Yu Cheng Ching-Yu Cheng Ching-Yu Cheng Smartphone-Acquired Anterior Segment Images for Deep Learning Prediction of Anterior Chamber Depth: A Proof-of-Concept Study Frontiers in Medicine primary angle-closure glaucoma glaucoma anterior chamber depth smartphone deep learning |
| title | Smartphone-Acquired Anterior Segment Images for Deep Learning Prediction of Anterior Chamber Depth: A Proof-of-Concept Study |
| title_full | Smartphone-Acquired Anterior Segment Images for Deep Learning Prediction of Anterior Chamber Depth: A Proof-of-Concept Study |
| title_fullStr | Smartphone-Acquired Anterior Segment Images for Deep Learning Prediction of Anterior Chamber Depth: A Proof-of-Concept Study |
| title_full_unstemmed | Smartphone-Acquired Anterior Segment Images for Deep Learning Prediction of Anterior Chamber Depth: A Proof-of-Concept Study |
| title_short | Smartphone-Acquired Anterior Segment Images for Deep Learning Prediction of Anterior Chamber Depth: A Proof-of-Concept Study |
| title_sort | smartphone acquired anterior segment images for deep learning prediction of anterior chamber depth a proof of concept study |
| topic | primary angle-closure glaucoma glaucoma anterior chamber depth smartphone deep learning |
| url | https://www.frontiersin.org/articles/10.3389/fmed.2022.912214/full |
| work_keys_str_mv | AT chaoxuqian smartphoneacquiredanteriorsegmentimagesfordeeplearningpredictionofanteriorchamberdepthaproofofconceptstudy AT chaoxuqian smartphoneacquiredanteriorsegmentimagesfordeeplearningpredictionofanteriorchamberdepthaproofofconceptstudy AT yixingjiang smartphoneacquiredanteriorsegmentimagesfordeeplearningpredictionofanteriorchamberdepthaproofofconceptstudy AT zhidasoh smartphoneacquiredanteriorsegmentimagesfordeeplearningpredictionofanteriorchamberdepthaproofofconceptstudy AT zhidasoh smartphoneacquiredanteriorsegmentimagesfordeeplearningpredictionofanteriorchamberdepthaproofofconceptstudy AT ganesansakthiselvam smartphoneacquiredanteriorsegmentimagesfordeeplearningpredictionofanteriorchamberdepthaproofofconceptstudy AT shuyuanxiao smartphoneacquiredanteriorsegmentimagesfordeeplearningpredictionofanteriorchamberdepthaproofofconceptstudy AT yihchungtham smartphoneacquiredanteriorsegmentimagesfordeeplearningpredictionofanteriorchamberdepthaproofofconceptstudy AT yihchungtham smartphoneacquiredanteriorsegmentimagesfordeeplearningpredictionofanteriorchamberdepthaproofofconceptstudy AT yihchungtham smartphoneacquiredanteriorsegmentimagesfordeeplearningpredictionofanteriorchamberdepthaproofofconceptstudy AT xinxingxu smartphoneacquiredanteriorsegmentimagesfordeeplearningpredictionofanteriorchamberdepthaproofofconceptstudy AT yongliu smartphoneacquiredanteriorsegmentimagesfordeeplearningpredictionofanteriorchamberdepthaproofofconceptstudy AT junli smartphoneacquiredanteriorsegmentimagesfordeeplearningpredictionofanteriorchamberdepthaproofofconceptstudy AT huazhong smartphoneacquiredanteriorsegmentimagesfordeeplearningpredictionofanteriorchamberdepthaproofofconceptstudy AT chingyucheng smartphoneacquiredanteriorsegmentimagesfordeeplearningpredictionofanteriorchamberdepthaproofofconceptstudy AT chingyucheng smartphoneacquiredanteriorsegmentimagesfordeeplearningpredictionofanteriorchamberdepthaproofofconceptstudy AT chingyucheng smartphoneacquiredanteriorsegmentimagesfordeeplearningpredictionofanteriorchamberdepthaproofofconceptstudy |