Search

Infectious keratitis, especially viral keratitis (VK), in resource-limited settings, can be a challenge to diagnose and carries a high risk of misdiagnosis contributing to significant ocular morbidity.employ and study the application of artificial intelligence-based deep learning (DL) algorithms to diagnose VK.A single-center retrospective study was conducted in a tertiary care center from January 2017 to December 2019 employing DL algorithm to diagnose VK from slit-lamp (SL) photographs. Three hundred and seven diffusely illuminated SL photographs from 285 eyes with polymerase chain reaction-proven herpes simplex viral stromal necrotizing keratitis (HSVNK) and culture-proven nonviral keratitis (NVK) were included. Patients having only HSV epithelial dendrites, endothelitis, mixed infection, and those with no SL photographs were excluded. DenseNet is a convolutional neural network, and the two main image datasets were divided into two subsets, one for training and the other for testing the algorithm. The performance of DenseNet was also compared with ResNet and Inception. Sensitivity, specificity, receiver operating characteristic (ROC) curve, and the area under the curve (AUC) were calculated.The accuracy of DenseNet on the test dataset was 72%, and it performed better than ResNet and Inception in the given task. The AUC for HSVNK was 0.73 with a sensitivity of 69.6% and specificity of 76.5%. The results were also validated using gradient-weighted class activation mapping (Grad-CAM), which successfully visualized the regions of input, which are significant for accurate predictions from these DL-based models.DL algorithm can be a positive aid to diagnose VK, especially in primary care centers where appropriate laboratory facilities or expert manpower are not available.

Aim: To compare the eyelid angle measured by using a manual method (Using protractor) and digital image analysis method (Using ImageJ software) at different distances of eye gaze. Methodology: This prospective cross-sectional study was conducted in the preclinical lab at Manipal College of Health Professions. Subjects with no eyelid abnormalities were included in the study and they were asked to fixate at different distances a) at 3 metre (Distance gaze) b) at 70 cm (Intermediategaze) and c) at 40 cm(Near gaze). Using a protractor, the eyelid angle measurements were repeated at various distances which comprised the manual measurement. In the image analysis method, images were captured during distance, intermediate and near gaze using smartphone placed on theside of the face. These images were then analysed using ImageJ software for determining eyelid angle using image analysis method. Palpebral fissure height, Palpebral fissure width, Interpupillary distance, Intercanthal width, Binocular width, Height of open upper lid were some additional anthropometry measurements that were done using meter scale and PD ruler. Results: The mean age of the participants was 20±0.5 years. Anthropometry measurements of the eyelid and Palpebral fissure were done using meter scaleand PD ruler. The mean and standard deviation of the measured parametersare as follows Interpupillary distance: 60.95±2.37 mm, Endo Inter canthal distance: 32.20±2.39 mm, Exo Inter cantal distance: 95.50±3.80 mm, Palpebralfissure height_OD: 12.11±1.32 mm, Palpebral fissure height_OS:12.16±1.46mm, PFW_OD: 32.00±1.10 mm, PFW_OS: 32.11±1.24 mm, Height of upper eyelids_OD: 10.26±1.66 mm and Height of upper eyelids_OS:10.42±1.83 mm. In the right eye, there was no statistically significant difference (p>0.05) between manual protractor method and digital image analysismethod at distance but there was a statistically significant difference (pConclusion: There is a significant difference in eyelid angle measured using manual protractor method and digital image analysis method. The measurement of eyelid angle serves as a critical reference point during cosmetic and reconstructive surgical interventions of the eyelid and accurate measurements are essential for preoperative assessment, surgical planning and postoperative evaluation. Hence more studies on the validation of the anthropometry measurements and eyelid angle using digital image analysis areessential to use digital image analysis in routine eye care practice.