A Convolutional Neural Network Using Anterior Segment Photos for Infectious Keratitis Identification
Vannarut Satitpitakul,1,2 Apiwit Puangsricharern,3 Surachet Yuktiratna,3 Yossapon Jaisarn,4 Keeratika Sangsao,4 Vilavun Puangsricharern,1,2 Ngamjit Kasetsuwan,1,2 Usanee Reinprayoon,1,2 Thanachaporn Kittipibul1,2 1Center of Excellence for Cornea and Stem Cell Transplantation, Department of Ophthalmo...
Saved in:
| Main Authors: | , , , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Dove Medical Press
2025-01-01
|
| Series: | Clinical Ophthalmology |
| Subjects: | |
| Online Access: | https://www.dovepress.com/a-convolutional-neural-network-using-anterior-segment-photos-for-infec-peer-reviewed-fulltext-article-OPTH |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1841555977887809536 |
|---|---|
| author | Satitpitakul V Puangsricharern A Yuktiratna S Jaisarn Y Sangsao K Puangsricharern V Kasetsuwan N Reinprayoon U Kittipibul T |
| author_facet | Satitpitakul V Puangsricharern A Yuktiratna S Jaisarn Y Sangsao K Puangsricharern V Kasetsuwan N Reinprayoon U Kittipibul T |
| author_sort | Satitpitakul V |
| collection | DOAJ |
| description | Vannarut Satitpitakul,1,2 Apiwit Puangsricharern,3 Surachet Yuktiratna,3 Yossapon Jaisarn,4 Keeratika Sangsao,4 Vilavun Puangsricharern,1,2 Ngamjit Kasetsuwan,1,2 Usanee Reinprayoon,1,2 Thanachaporn Kittipibul1,2 1Center of Excellence for Cornea and Stem Cell Transplantation, Department of Ophthalmology, Faculty of Medicine, Chulalongkorn University, Bangkok, Thailand; 2Excellence Center for Cornea and Stem Cell Transplantation, Department of Ophthalmology, King Chulalongkorn Memorial Hospital, Thai Red Cross Society, Bangkok, Thailand; 3IM Impower Company Limited, Bangkok, Thailand; 4Faculty of Medicine, Chulalongkorn University, Bangkok, ThailandCorrespondence: Vannarut Satitpitakul, Department of Ophthalmology, King Chulalongkorn Memorial Hospital, 1873 Rama 4 Road, Pathumwan, Bangkok, 10330, Thailand, Tel +66-894959022, Email Vannarut.S@chula.ac.thPurpose: To develop a comprehensively deep learning algorithm to differentiate between bacterial keratitis, fungal keratitis, non-infectious corneal lesions, and normal corneas.Methods: This retrospective study collected slit-lamp photos of patients with bacterial keratitis, fungal keratitis, non-infectious corneal lesions, and normal cornea. Causative organisms of infectious keratitis were identified by either positive culture or clinical response to single treatment. Convolutional neural networks (ResNet50, DenseNet121, VGG19) and Ensemble with probability weighting were used to develop a deep learning algorithm. The performance including accuracy, precision, recall, F1 score, specificity and AUC has been reported.Results: Total of 6478 photos from 2171 eyes, composed of 2400 bacterial keratitis, 1616 fungal keratitis, 1545 non-infectious corneal lesions, and 917 normal corneas were collected from hospital database. DenseNet121 demonstrated the best performance among three convolutional neural networks with the accuracy of 0.8 (95% CI 0.74– 0.86). The ensemble technique showed higher performance than single algorithm with the accuracy of 0.83 (95% 0.78– 0.88).Conclusion: Convolutional neural networks with ensemble techniques provided the best performance in discriminating bacterial keratitis, fungal keratitis, non-infectious corneal lesions, and normal corneas. Our models can be used as a screening tool for non-ophthalmic health care providers and ophthalmologists for rapid provisional diagnosis of infectious keratitis.Keywords: infectious keratitis, cornea ulcer, keratitis, conventional neural network, deep learning algorithm |
| format | Article |
| id | doaj-art-66df10a440ae4445807391ffe3841364 |
| institution | Kabale University |
| issn | 1177-5483 |
| language | English |
| publishDate | 2025-01-01 |
| publisher | Dove Medical Press |
| record_format | Article |
| series | Clinical Ophthalmology |
| spelling | doaj-art-66df10a440ae4445807391ffe38413642025-01-07T16:42:40ZengDove Medical PressClinical Ophthalmology1177-54832025-01-01Volume 19738199039A Convolutional Neural Network Using Anterior Segment Photos for Infectious Keratitis IdentificationSatitpitakul VPuangsricharern AYuktiratna SJaisarn YSangsao KPuangsricharern VKasetsuwan NReinprayoon UKittipibul TVannarut Satitpitakul,1,2 Apiwit Puangsricharern,3 Surachet Yuktiratna,3 Yossapon Jaisarn,4 Keeratika Sangsao,4 Vilavun Puangsricharern,1,2 Ngamjit Kasetsuwan,1,2 Usanee Reinprayoon,1,2 Thanachaporn Kittipibul1,2 1Center of Excellence for Cornea and Stem Cell Transplantation, Department of Ophthalmology, Faculty of Medicine, Chulalongkorn University, Bangkok, Thailand; 2Excellence Center for Cornea and Stem Cell Transplantation, Department of Ophthalmology, King Chulalongkorn Memorial Hospital, Thai Red Cross Society, Bangkok, Thailand; 3IM Impower Company Limited, Bangkok, Thailand; 4Faculty of Medicine, Chulalongkorn University, Bangkok, ThailandCorrespondence: Vannarut Satitpitakul, Department of Ophthalmology, King Chulalongkorn Memorial Hospital, 1873 Rama 4 Road, Pathumwan, Bangkok, 10330, Thailand, Tel +66-894959022, Email Vannarut.S@chula.ac.thPurpose: To develop a comprehensively deep learning algorithm to differentiate between bacterial keratitis, fungal keratitis, non-infectious corneal lesions, and normal corneas.Methods: This retrospective study collected slit-lamp photos of patients with bacterial keratitis, fungal keratitis, non-infectious corneal lesions, and normal cornea. Causative organisms of infectious keratitis were identified by either positive culture or clinical response to single treatment. Convolutional neural networks (ResNet50, DenseNet121, VGG19) and Ensemble with probability weighting were used to develop a deep learning algorithm. The performance including accuracy, precision, recall, F1 score, specificity and AUC has been reported.Results: Total of 6478 photos from 2171 eyes, composed of 2400 bacterial keratitis, 1616 fungal keratitis, 1545 non-infectious corneal lesions, and 917 normal corneas were collected from hospital database. DenseNet121 demonstrated the best performance among three convolutional neural networks with the accuracy of 0.8 (95% CI 0.74– 0.86). The ensemble technique showed higher performance than single algorithm with the accuracy of 0.83 (95% 0.78– 0.88).Conclusion: Convolutional neural networks with ensemble techniques provided the best performance in discriminating bacterial keratitis, fungal keratitis, non-infectious corneal lesions, and normal corneas. Our models can be used as a screening tool for non-ophthalmic health care providers and ophthalmologists for rapid provisional diagnosis of infectious keratitis.Keywords: infectious keratitis, cornea ulcer, keratitis, conventional neural network, deep learning algorithmhttps://www.dovepress.com/a-convolutional-neural-network-using-anterior-segment-photos-for-infec-peer-reviewed-fulltext-article-OPTHinfectious keratitiscornea ulcerkeratitisconventional neural networkdeep learning algorithm |
| spellingShingle | Satitpitakul V Puangsricharern A Yuktiratna S Jaisarn Y Sangsao K Puangsricharern V Kasetsuwan N Reinprayoon U Kittipibul T A Convolutional Neural Network Using Anterior Segment Photos for Infectious Keratitis Identification Clinical Ophthalmology infectious keratitis cornea ulcer keratitis conventional neural network deep learning algorithm |
| title | A Convolutional Neural Network Using Anterior Segment Photos for Infectious Keratitis Identification |
| title_full | A Convolutional Neural Network Using Anterior Segment Photos for Infectious Keratitis Identification |
| title_fullStr | A Convolutional Neural Network Using Anterior Segment Photos for Infectious Keratitis Identification |
| title_full_unstemmed | A Convolutional Neural Network Using Anterior Segment Photos for Infectious Keratitis Identification |
| title_short | A Convolutional Neural Network Using Anterior Segment Photos for Infectious Keratitis Identification |
| title_sort | convolutional neural network using anterior segment photos for infectious keratitis identification |
| topic | infectious keratitis cornea ulcer keratitis conventional neural network deep learning algorithm |
| url | https://www.dovepress.com/a-convolutional-neural-network-using-anterior-segment-photos-for-infec-peer-reviewed-fulltext-article-OPTH |
| work_keys_str_mv | AT satitpitakulv aconvolutionalneuralnetworkusinganteriorsegmentphotosforinfectiouskeratitisidentification AT puangsricharerna aconvolutionalneuralnetworkusinganteriorsegmentphotosforinfectiouskeratitisidentification AT yuktiratnas aconvolutionalneuralnetworkusinganteriorsegmentphotosforinfectiouskeratitisidentification AT jaisarny aconvolutionalneuralnetworkusinganteriorsegmentphotosforinfectiouskeratitisidentification AT sangsaok aconvolutionalneuralnetworkusinganteriorsegmentphotosforinfectiouskeratitisidentification AT puangsricharernv aconvolutionalneuralnetworkusinganteriorsegmentphotosforinfectiouskeratitisidentification AT kasetsuwann aconvolutionalneuralnetworkusinganteriorsegmentphotosforinfectiouskeratitisidentification AT reinprayoonu aconvolutionalneuralnetworkusinganteriorsegmentphotosforinfectiouskeratitisidentification AT kittipibult aconvolutionalneuralnetworkusinganteriorsegmentphotosforinfectiouskeratitisidentification AT satitpitakulv convolutionalneuralnetworkusinganteriorsegmentphotosforinfectiouskeratitisidentification AT puangsricharerna convolutionalneuralnetworkusinganteriorsegmentphotosforinfectiouskeratitisidentification AT yuktiratnas convolutionalneuralnetworkusinganteriorsegmentphotosforinfectiouskeratitisidentification AT jaisarny convolutionalneuralnetworkusinganteriorsegmentphotosforinfectiouskeratitisidentification AT sangsaok convolutionalneuralnetworkusinganteriorsegmentphotosforinfectiouskeratitisidentification AT puangsricharernv convolutionalneuralnetworkusinganteriorsegmentphotosforinfectiouskeratitisidentification AT kasetsuwann convolutionalneuralnetworkusinganteriorsegmentphotosforinfectiouskeratitisidentification AT reinprayoonu convolutionalneuralnetworkusinganteriorsegmentphotosforinfectiouskeratitisidentification AT kittipibult convolutionalneuralnetworkusinganteriorsegmentphotosforinfectiouskeratitisidentification |