Use of explainable AI on slit-lamp images of anterior surface of eyes to diagnose allergic conjunctival diseases
Background: Artificial intelligence (AI) is a promising new technology that has the potential of diagnosing allergic conjunctival diseases (ACDs). However, its development is slowed by the absence of a tailored image database and explainable AI models. Thus, the purpose of this study was to develop...
Saved in:
| Main Authors: | , , , , , , , , , , , , , , , , , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Elsevier
2025-01-01
|
| Series: | Allergology International |
| Subjects: | |
| Online Access: | http://www.sciencedirect.com/science/article/pii/S1323893024000777 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1850076828235988992 |
|---|---|
| author | Michiko Yonehara Yuji Nakagawa Yuji Ayatsuka Yuko Hara Jun Shoji Nobuyuki Ebihara Takenori Inomata Tianxiang Huang Ken Nagino Ken Fukuda Tatsuma Kishimoto Tamaki Sumi Atsuki Fukushima Hiroshi Fujishima Moeko Kawai Etsuko Takamura Eiichi Uchio Kenichi Namba Ayumi Koyama Tomoko Haruki Shin-ich Sasaki Yumiko Shimizu Dai Miyazaki |
| author_facet | Michiko Yonehara Yuji Nakagawa Yuji Ayatsuka Yuko Hara Jun Shoji Nobuyuki Ebihara Takenori Inomata Tianxiang Huang Ken Nagino Ken Fukuda Tatsuma Kishimoto Tamaki Sumi Atsuki Fukushima Hiroshi Fujishima Moeko Kawai Etsuko Takamura Eiichi Uchio Kenichi Namba Ayumi Koyama Tomoko Haruki Shin-ich Sasaki Yumiko Shimizu Dai Miyazaki |
| author_sort | Michiko Yonehara |
| collection | DOAJ |
| description | Background: Artificial intelligence (AI) is a promising new technology that has the potential of diagnosing allergic conjunctival diseases (ACDs). However, its development is slowed by the absence of a tailored image database and explainable AI models. Thus, the purpose of this study was to develop an explainable AI model that can not only diagnose ACDs but also present the basis for the diagnosis. Methods: A dataset of 4942 slit-lamp images from 10 ophthalmological institutions across Japan were used as the image database. A sequential pipeline of segmentation AI was constructed to identify 12 clinical findings in 1038 images of seasonal and perennial allergic conjunctivitis (AC), atopic keratoconjunctivitis (AKC), vernal keratoconjunctivitis (VKC), giant papillary conjunctivitis (GPC), and normal subjects. The performance of the pipeline was evaluated by determining its ability to obtain explainable results through the extraction of the findings. Its diagnostic accuracy was determined for 4 severity-based diagnosis classification of AC, AKC/VKC, GPC, and normal. Results: Segmentation AI pipeline efficiently extracted crucial ACD indicators including conjunctival hyperemia, giant papillae, and shield ulcer, and offered interpretable insights. The AI pipeline diagnosis had a high diagnostic accuracy of 86.2%, and that of the board-certified ophthalmologists was 60.0%. The pipeline had a high classification performance, and the area under the curve (AUC) was 0.959 for AC, 0.905 for normal subjects, 0.847 for GPC, 0.829 for VKC, and 0.790 for AKC. Conclusions: An explainable AI model created by a comprehensive image database can be used for diagnosing ACDs with high degree of accuracy. |
| format | Article |
| id | doaj-art-368cf7e9398a4c2192af485d14fdce9a |
| institution | DOAJ |
| issn | 1323-8930 |
| language | English |
| publishDate | 2025-01-01 |
| publisher | Elsevier |
| record_format | Article |
| series | Allergology International |
| spelling | doaj-art-368cf7e9398a4c2192af485d14fdce9a2025-08-20T02:45:56ZengElsevierAllergology International1323-89302025-01-01741869610.1016/j.alit.2024.07.004Use of explainable AI on slit-lamp images of anterior surface of eyes to diagnose allergic conjunctival diseasesMichiko Yonehara0Yuji Nakagawa1Yuji Ayatsuka2Yuko Hara3Jun Shoji4Nobuyuki Ebihara5Takenori Inomata6Tianxiang Huang7Ken Nagino8Ken Fukuda9Tatsuma Kishimoto10Tamaki Sumi11Atsuki Fukushima12Hiroshi Fujishima13Moeko Kawai14Etsuko Takamura15Eiichi Uchio16Kenichi Namba17Ayumi Koyama18Tomoko Haruki19Shin-ich Sasaki20Yumiko Shimizu21Dai Miyazaki22Division of Ophthalmology and Visual Science, Faculty of Medicine, Tottori University, Tottori, JapanTechnology Laboratory, Cresco Ltd., Tokyo, JapanTechnology Laboratory, Cresco Ltd., Tokyo, JapanDepartment of Ophthalmology, Ehime University Graduate School of Medicine, Ehime, JapanDivision of Ophthalmology, Department of Visual Sciences, Nihon University School of Medicine, Tokyo, JapanDepartment of Ophthalmology, Juntendo University Urayasu Hospital, Chiba, JapanDepartment of Ophthalmology, Juntendo University Graduate School of Medicine, Tokyo, Japan; Department of Hospital Administration, Juntendo University Graduate School of Medicine, Tokyo, Japan; AI Incubation Farm, Juntendo University Graduate School of Medicine, Tokyo, JapanDepartment of Ophthalmology, Juntendo University Graduate School of Medicine, Tokyo, JapanDepartment of Ophthalmology, Juntendo University Graduate School of Medicine, Tokyo, Japan; Department of Hospital Administration, Juntendo University Graduate School of Medicine, Tokyo, JapanDepartment of Ophthalmology, Kochi Medical School, Kochi University, Kochi, JapanDepartment of Ophthalmology, Kochi Medical School, Kochi University, Kochi, JapanDepartment of Ophthalmology, Kochi Medical School, Kochi University, Kochi, JapanDepartment of Ophthalmology, Tsukazaki Hospital, Hyogo, JapanDepartment of Ophthalmology, School of Dental Medicine, Tsurumi University, Kanagawa, JapanDepartment of Ophthalmology, School of Medicine, Tokyo Women's Medical University, Tokyo, JapanDepartment of Ophthalmology, School of Medicine, Tokyo Women's Medical University, Tokyo, JapanDepartment of Ophthalmology, Faculty of Medicine, Fukuoka University, Fukuoka, JapanDepartment of Ophthalmology, Faculty of Medicine, Graduate School of Medicine, Hokkaido University, Sapporo, JapanDivision of Ophthalmology and Visual Science, Faculty of Medicine, Tottori University, Tottori, JapanDivision of Ophthalmology and Visual Science, Faculty of Medicine, Tottori University, Tottori, JapanDivision of Ophthalmology and Visual Science, Faculty of Medicine, Tottori University, Tottori, JapanDivision of Ophthalmology and Visual Science, Faculty of Medicine, Tottori University, Tottori, JapanDivision of Ophthalmology and Visual Science, Faculty of Medicine, Tottori University, Tottori, Japan; Corresponding author. Division of Ophthalmology and Visual Science, Faculty of Medicine, Tottori University, 36-1 Nishi-cho, Yonago, Tottori 683-8504, Japan.Background: Artificial intelligence (AI) is a promising new technology that has the potential of diagnosing allergic conjunctival diseases (ACDs). However, its development is slowed by the absence of a tailored image database and explainable AI models. Thus, the purpose of this study was to develop an explainable AI model that can not only diagnose ACDs but also present the basis for the diagnosis. Methods: A dataset of 4942 slit-lamp images from 10 ophthalmological institutions across Japan were used as the image database. A sequential pipeline of segmentation AI was constructed to identify 12 clinical findings in 1038 images of seasonal and perennial allergic conjunctivitis (AC), atopic keratoconjunctivitis (AKC), vernal keratoconjunctivitis (VKC), giant papillary conjunctivitis (GPC), and normal subjects. The performance of the pipeline was evaluated by determining its ability to obtain explainable results through the extraction of the findings. Its diagnostic accuracy was determined for 4 severity-based diagnosis classification of AC, AKC/VKC, GPC, and normal. Results: Segmentation AI pipeline efficiently extracted crucial ACD indicators including conjunctival hyperemia, giant papillae, and shield ulcer, and offered interpretable insights. The AI pipeline diagnosis had a high diagnostic accuracy of 86.2%, and that of the board-certified ophthalmologists was 60.0%. The pipeline had a high classification performance, and the area under the curve (AUC) was 0.959 for AC, 0.905 for normal subjects, 0.847 for GPC, 0.829 for VKC, and 0.790 for AKC. Conclusions: An explainable AI model created by a comprehensive image database can be used for diagnosing ACDs with high degree of accuracy.http://www.sciencedirect.com/science/article/pii/S1323893024000777Allergic conjunctivitisArtificial intelligenceAtopic keratoconjunctivitisDeep learningVernal keratoconjunctivitis |
| spellingShingle | Michiko Yonehara Yuji Nakagawa Yuji Ayatsuka Yuko Hara Jun Shoji Nobuyuki Ebihara Takenori Inomata Tianxiang Huang Ken Nagino Ken Fukuda Tatsuma Kishimoto Tamaki Sumi Atsuki Fukushima Hiroshi Fujishima Moeko Kawai Etsuko Takamura Eiichi Uchio Kenichi Namba Ayumi Koyama Tomoko Haruki Shin-ich Sasaki Yumiko Shimizu Dai Miyazaki Use of explainable AI on slit-lamp images of anterior surface of eyes to diagnose allergic conjunctival diseases Allergology International Allergic conjunctivitis Artificial intelligence Atopic keratoconjunctivitis Deep learning Vernal keratoconjunctivitis |
| title | Use of explainable AI on slit-lamp images of anterior surface of eyes to diagnose allergic conjunctival diseases |
| title_full | Use of explainable AI on slit-lamp images of anterior surface of eyes to diagnose allergic conjunctival diseases |
| title_fullStr | Use of explainable AI on slit-lamp images of anterior surface of eyes to diagnose allergic conjunctival diseases |
| title_full_unstemmed | Use of explainable AI on slit-lamp images of anterior surface of eyes to diagnose allergic conjunctival diseases |
| title_short | Use of explainable AI on slit-lamp images of anterior surface of eyes to diagnose allergic conjunctival diseases |
| title_sort | use of explainable ai on slit lamp images of anterior surface of eyes to diagnose allergic conjunctival diseases |
| topic | Allergic conjunctivitis Artificial intelligence Atopic keratoconjunctivitis Deep learning Vernal keratoconjunctivitis |
| url | http://www.sciencedirect.com/science/article/pii/S1323893024000777 |
| work_keys_str_mv | AT michikoyonehara useofexplainableaionslitlampimagesofanteriorsurfaceofeyestodiagnoseallergicconjunctivaldiseases AT yujinakagawa useofexplainableaionslitlampimagesofanteriorsurfaceofeyestodiagnoseallergicconjunctivaldiseases AT yujiayatsuka useofexplainableaionslitlampimagesofanteriorsurfaceofeyestodiagnoseallergicconjunctivaldiseases AT yukohara useofexplainableaionslitlampimagesofanteriorsurfaceofeyestodiagnoseallergicconjunctivaldiseases AT junshoji useofexplainableaionslitlampimagesofanteriorsurfaceofeyestodiagnoseallergicconjunctivaldiseases AT nobuyukiebihara useofexplainableaionslitlampimagesofanteriorsurfaceofeyestodiagnoseallergicconjunctivaldiseases AT takenoriinomata useofexplainableaionslitlampimagesofanteriorsurfaceofeyestodiagnoseallergicconjunctivaldiseases AT tianxianghuang useofexplainableaionslitlampimagesofanteriorsurfaceofeyestodiagnoseallergicconjunctivaldiseases AT kennagino useofexplainableaionslitlampimagesofanteriorsurfaceofeyestodiagnoseallergicconjunctivaldiseases AT kenfukuda useofexplainableaionslitlampimagesofanteriorsurfaceofeyestodiagnoseallergicconjunctivaldiseases AT tatsumakishimoto useofexplainableaionslitlampimagesofanteriorsurfaceofeyestodiagnoseallergicconjunctivaldiseases AT tamakisumi useofexplainableaionslitlampimagesofanteriorsurfaceofeyestodiagnoseallergicconjunctivaldiseases AT atsukifukushima useofexplainableaionslitlampimagesofanteriorsurfaceofeyestodiagnoseallergicconjunctivaldiseases AT hiroshifujishima useofexplainableaionslitlampimagesofanteriorsurfaceofeyestodiagnoseallergicconjunctivaldiseases AT moekokawai useofexplainableaionslitlampimagesofanteriorsurfaceofeyestodiagnoseallergicconjunctivaldiseases AT etsukotakamura useofexplainableaionslitlampimagesofanteriorsurfaceofeyestodiagnoseallergicconjunctivaldiseases AT eiichiuchio useofexplainableaionslitlampimagesofanteriorsurfaceofeyestodiagnoseallergicconjunctivaldiseases AT kenichinamba useofexplainableaionslitlampimagesofanteriorsurfaceofeyestodiagnoseallergicconjunctivaldiseases AT ayumikoyama useofexplainableaionslitlampimagesofanteriorsurfaceofeyestodiagnoseallergicconjunctivaldiseases AT tomokoharuki useofexplainableaionslitlampimagesofanteriorsurfaceofeyestodiagnoseallergicconjunctivaldiseases AT shinichsasaki useofexplainableaionslitlampimagesofanteriorsurfaceofeyestodiagnoseallergicconjunctivaldiseases AT yumikoshimizu useofexplainableaionslitlampimagesofanteriorsurfaceofeyestodiagnoseallergicconjunctivaldiseases AT daimiyazaki useofexplainableaionslitlampimagesofanteriorsurfaceofeyestodiagnoseallergicconjunctivaldiseases |