Prediction of the ectasia screening index from raw Casia2 volume data for keratoconus identification by using convolutional neural networks

Purpose Prediction of Ectasia Screening Index (ESI), an estimator provided by the Casia2 for identifying keratoconus, from raw Optical Coherence Tomography (OCT) data with Convolutional Neural Networks (CNN). Methods Three CNN architectures (ResNet18, DenseNet121 and EfficientNetB0) were employed to predict the ESI. Mean Absolute Error (MAE) was used as the performance metric for predicting the ESI by the adapted CNN models on the test set. Scans with an ESI value higher than a certain threshold were classified as Keratoconus, while the remaining scans were classified as Not Keratoconus. The models’ performance was evaluated using metrics such as accuracy, sensitivity, specificity, Positive Predictive Value (PPV) and F1 score on data collected from patients examined at the eye clinic of the Homburg University Hospital. The raw data from the Casia2 device, in 3dv format, was converted into 16 images per examination of one eye. For the training, validation and testing phases, 3689, 1050 and 1078 scans (3dv files) were selected, respectively. Results In the prediction of the ESI, the MAE values for the adapted ResNet18, DenseNet121 and EfficientNetB0, rounded to two decimal places, were 7.15, 6.64 and 5.86, respectively. In the classification task, the three networks yielded an accuracy of 94.80%, 95.27% and 95.83%, respectively; a sensitivity of 92.07%, 94.64% and 94.17%, respectively; a specificity of 96.61%, 95.69% and 96.92%, respectively; a PPV of 94.72%, 93.55% and 95.28%, respectively; and a F1 score of 93.38%, 94.09% and 94.72%, respectively. Conclusions Our results show that the prediction of keratokonus based on the ESI values estimated from raw data outperforms previous approaches using processed data. Adapted EfficientNetB0 outperformed both the other adapted models and those in state-of-the-art studies, with the highest accuracy and F1 score.