Using Slit-Lamp Images for Deep Learning-Based Identification of Bacterial and Fungal Keratitis: Model Development and Validation with Different Convolutional Neural Networks
In this study, we aimed to develop a deep learning model for identifying bacterial keratitis (BK) and fungal keratitis (FK) by using slit-lamp images. We retrospectively collected slit-lamp images of patients with culture-proven microbial keratitis between 1 January 2010 and 31 December 2019 from tw...
| Main Authors: | , , , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2021-07-01
|
| Series: | Diagnostics |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2075-4418/11/7/1246 |
| _version_ | 1797527306246291456 |
|---|---|
| author | Ning Hung Andy Kuan-Yu Shih Chihung Lin Ming-Tse Kuo Yih-Shiou Hwang Wei-Chi Wu Chang-Fu Kuo Eugene Yu-Chuan Kang Ching-Hsi Hsiao |
| author_facet | Ning Hung Andy Kuan-Yu Shih Chihung Lin Ming-Tse Kuo Yih-Shiou Hwang Wei-Chi Wu Chang-Fu Kuo Eugene Yu-Chuan Kang Ching-Hsi Hsiao |
| author_sort | Ning Hung |
| collection | DOAJ |
| description | In this study, we aimed to develop a deep learning model for identifying bacterial keratitis (BK) and fungal keratitis (FK) by using slit-lamp images. We retrospectively collected slit-lamp images of patients with culture-proven microbial keratitis between 1 January 2010 and 31 December 2019 from two medical centers in Taiwan. We constructed a deep learning algorithm consisting of a segmentation model for cropping cornea images and a classification model that applies different convolutional neural networks (CNNs) to differentiate between FK and BK. The CNNs included DenseNet121, DenseNet161, DenseNet169, DenseNet201, EfficientNetB3, InceptionV3, ResNet101, and ResNet50. The model performance was evaluated and presented as the area under the curve (AUC) of the receiver operating characteristic curves. A gradient-weighted class activation mapping technique was used to plot the heat map of the model. By using 1330 images from 580 patients, the deep learning algorithm achieved the highest average accuracy of 80.0%. Using different CNNs, the diagnostic accuracy for BK ranged from 79.6% to 95.9%, and that for FK ranged from 26.3% to 65.8%. The CNN of DenseNet161 showed the best model performance, with an AUC of 0.85 for both BK and FK. The heat maps revealed that the model was able to identify the corneal infiltrations. The model showed a better diagnostic accuracy than the previously reported diagnostic performance of both general ophthalmologists and corneal specialists. |
| first_indexed | 2024-03-10T09:42:05Z |
| format | Article |
| id | doaj.art-5a02d1f6a33e4f149613cfa9614d0229 |
| institution | Directory Open Access Journal |
| issn | 2075-4418 |
| language | English |
| last_indexed | 2024-03-10T09:42:05Z |
| publishDate | 2021-07-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Diagnostics |
| spelling | doaj.art-5a02d1f6a33e4f149613cfa9614d02292023-11-22T03:34:57ZengMDPI AGDiagnostics2075-44182021-07-01117124610.3390/diagnostics11071246Using Slit-Lamp Images for Deep Learning-Based Identification of Bacterial and Fungal Keratitis: Model Development and Validation with Different Convolutional Neural NetworksNing Hung0Andy Kuan-Yu Shih1Chihung Lin2Ming-Tse Kuo3Yih-Shiou Hwang4Wei-Chi Wu5Chang-Fu Kuo6Eugene Yu-Chuan Kang7Ching-Hsi Hsiao8Department of Ophthalmology, Chang Gung Memorial Hospital, Linkou Medical Center, No. 5 Fu-Hsin Rd, Kweishan, Taoyuan 333, TaiwanCenter for Artificial Intelligence in Medicine, Chang Gung Memorial Hospital, Linkou Medical Center, No. 5 Fu-Hsin Rd, Kweishan, Taoyuan 333, TaiwanCenter for Artificial Intelligence in Medicine, Chang Gung Memorial Hospital, Linkou Medical Center, No. 5 Fu-Hsin Rd, Kweishan, Taoyuan 333, TaiwanDepartment of Ophthalmology, Kaohsiung Chang Gung Memorial Hospital, No. 123, Dapi Rd, Niaosong, Kaohsiung 833, TaiwanDepartment of Ophthalmology, Chang Gung Memorial Hospital, Linkou Medical Center, No. 5 Fu-Hsin Rd, Kweishan, Taoyuan 333, TaiwanDepartment of Ophthalmology, Chang Gung Memorial Hospital, Linkou Medical Center, No. 5 Fu-Hsin Rd, Kweishan, Taoyuan 333, TaiwanCenter for Artificial Intelligence in Medicine, Chang Gung Memorial Hospital, Linkou Medical Center, No. 5 Fu-Hsin Rd, Kweishan, Taoyuan 333, TaiwanDepartment of Ophthalmology, Chang Gung Memorial Hospital, Linkou Medical Center, No. 5 Fu-Hsin Rd, Kweishan, Taoyuan 333, TaiwanDepartment of Ophthalmology, Chang Gung Memorial Hospital, Linkou Medical Center, No. 5 Fu-Hsin Rd, Kweishan, Taoyuan 333, TaiwanIn this study, we aimed to develop a deep learning model for identifying bacterial keratitis (BK) and fungal keratitis (FK) by using slit-lamp images. We retrospectively collected slit-lamp images of patients with culture-proven microbial keratitis between 1 January 2010 and 31 December 2019 from two medical centers in Taiwan. We constructed a deep learning algorithm consisting of a segmentation model for cropping cornea images and a classification model that applies different convolutional neural networks (CNNs) to differentiate between FK and BK. The CNNs included DenseNet121, DenseNet161, DenseNet169, DenseNet201, EfficientNetB3, InceptionV3, ResNet101, and ResNet50. The model performance was evaluated and presented as the area under the curve (AUC) of the receiver operating characteristic curves. A gradient-weighted class activation mapping technique was used to plot the heat map of the model. By using 1330 images from 580 patients, the deep learning algorithm achieved the highest average accuracy of 80.0%. Using different CNNs, the diagnostic accuracy for BK ranged from 79.6% to 95.9%, and that for FK ranged from 26.3% to 65.8%. The CNN of DenseNet161 showed the best model performance, with an AUC of 0.85 for both BK and FK. The heat maps revealed that the model was able to identify the corneal infiltrations. The model showed a better diagnostic accuracy than the previously reported diagnostic performance of both general ophthalmologists and corneal specialists.https://www.mdpi.com/2075-4418/11/7/1246deep learninginfectious keratitiscropped corneal imageslit-lamp images |
| spellingShingle | Ning Hung Andy Kuan-Yu Shih Chihung Lin Ming-Tse Kuo Yih-Shiou Hwang Wei-Chi Wu Chang-Fu Kuo Eugene Yu-Chuan Kang Ching-Hsi Hsiao Using Slit-Lamp Images for Deep Learning-Based Identification of Bacterial and Fungal Keratitis: Model Development and Validation with Different Convolutional Neural Networks Diagnostics deep learning infectious keratitis cropped corneal image slit-lamp images |
| title | Using Slit-Lamp Images for Deep Learning-Based Identification of Bacterial and Fungal Keratitis: Model Development and Validation with Different Convolutional Neural Networks |
| title_full | Using Slit-Lamp Images for Deep Learning-Based Identification of Bacterial and Fungal Keratitis: Model Development and Validation with Different Convolutional Neural Networks |
| title_fullStr | Using Slit-Lamp Images for Deep Learning-Based Identification of Bacterial and Fungal Keratitis: Model Development and Validation with Different Convolutional Neural Networks |
| title_full_unstemmed | Using Slit-Lamp Images for Deep Learning-Based Identification of Bacterial and Fungal Keratitis: Model Development and Validation with Different Convolutional Neural Networks |
| title_short | Using Slit-Lamp Images for Deep Learning-Based Identification of Bacterial and Fungal Keratitis: Model Development and Validation with Different Convolutional Neural Networks |
| title_sort | using slit lamp images for deep learning based identification of bacterial and fungal keratitis model development and validation with different convolutional neural networks |
| topic | deep learning infectious keratitis cropped corneal image slit-lamp images |
| url | https://www.mdpi.com/2075-4418/11/7/1246 |
| work_keys_str_mv | AT ninghung usingslitlampimagesfordeeplearningbasedidentificationofbacterialandfungalkeratitismodeldevelopmentandvalidationwithdifferentconvolutionalneuralnetworks AT andykuanyushih usingslitlampimagesfordeeplearningbasedidentificationofbacterialandfungalkeratitismodeldevelopmentandvalidationwithdifferentconvolutionalneuralnetworks AT chihunglin usingslitlampimagesfordeeplearningbasedidentificationofbacterialandfungalkeratitismodeldevelopmentandvalidationwithdifferentconvolutionalneuralnetworks AT mingtsekuo usingslitlampimagesfordeeplearningbasedidentificationofbacterialandfungalkeratitismodeldevelopmentandvalidationwithdifferentconvolutionalneuralnetworks AT yihshiouhwang usingslitlampimagesfordeeplearningbasedidentificationofbacterialandfungalkeratitismodeldevelopmentandvalidationwithdifferentconvolutionalneuralnetworks AT weichiwu usingslitlampimagesfordeeplearningbasedidentificationofbacterialandfungalkeratitismodeldevelopmentandvalidationwithdifferentconvolutionalneuralnetworks AT changfukuo usingslitlampimagesfordeeplearningbasedidentificationofbacterialandfungalkeratitismodeldevelopmentandvalidationwithdifferentconvolutionalneuralnetworks AT eugeneyuchuankang usingslitlampimagesfordeeplearningbasedidentificationofbacterialandfungalkeratitismodeldevelopmentandvalidationwithdifferentconvolutionalneuralnetworks AT chinghsihsiao usingslitlampimagesfordeeplearningbasedidentificationofbacterialandfungalkeratitismodeldevelopmentandvalidationwithdifferentconvolutionalneuralnetworks |