Background - Non-pulmonary vein (NPV) trigger has been reported as an important predictor of recurrence post-atrial fibrillation (AF) ablation. Elimination of NPV triggers can reduce the recurrence of post-ablation AF. Deep learning was applied to pre-ablation pulmonary vein computed tomography (PVCT) geometric slices to create a prediction model for NPV triggers in patients with paroxysmal atrial fibrillation (PAF). Methods - We retrospectively analyzed 521 PAF patients who underwent catheter ablation of PAF. Among them, PVCT geometric slices from 358 non-recurrent AF patients (1-3 mm interspace per slice, 20-200 slices for each patient, ranging from the upper border of the left atrium to the bottom of the heart, for a total of 23683 images of slices) were used in the deep learning process, the ResNet34 of the neural network, to create the prediction model of the NPV trigger. There were 298 (83.2%) patients with only pulmonary vein (PV) triggers and 60 (16.8%) patients with NPV triggers +/- PV triggers. The patients were randomly assigned to either training, validation or test groups and their data was allocated according to those sets. The image datasets were split into training (n=17340), validation (n=3491), and testing (n=2852) groups, which had completely independent sets of patients. Results - The accuracy of prediction in each PVCT image for NPV trigger was up to 82.4±2.0%. The sensitivity and specificity were 64.3±5.4% and 88.4±1.9%, respectively. For each patient, the accuracy of prediction for a NPV trigger was 88.6±2.3%. The sensitivity and specificity were 75.0±5.8% and 95.7±1.8%, respectively. The area under the curve (AUC) for each image and patient were 0.82±0.01 and 0.88±0.07, respectively. Conclusions - The deep learning model using pre-ablation PVCT can be applied to predict the trigger origins in PAF patients receiving catheter ablation. The application of this model may identify patients with a high risk of NPV trigger before ablation.
Automated glaucoma detection using deep learning may increase the diagnostic rate of glaucoma to prevent blindness, but generalizable models are currently unavailable despite the use of huge training datasets. This study aims to evaluate the performance of a convolutional neural network (CNN) classifier trained with a limited number of high-quality fundus images in detecting glaucoma and methods to improve its performance across different datasets. A CNN classifier was constructed using EfficientNet B3 and 944 images collected from one medical center (core model) and externally validated using three datasets. The performance of the core model was compared with (1) the integrated model constructed by using all training images from the four datasets and (2) the dataset-specific model built by fine-tuning the core model with training images from the external datasets. The diagnostic accuracy of the core model was 95.62% but dropped to ranges of 52.5–80.0% on the external datasets. Dataset-specific models exhibited superior diagnostic performance on the external datasets compared to other models, with a diagnostic accuracy of 87.50–92.5%. The findings suggest that dataset-specific tuning of the core CNN classifier effectively improves its applicability across different datasets when increasing training images fails to achieve generalization.