Search

Purpose: To evaluate graft detachment after Descemet membrane endothelial keratoplasty (DMEK) in pseudophakic eyes and DMEK combined with cataract surgery (triple DMEK). Design: Analysis of 3 single-center prospective cohort studies and 1 randomized controlled trial. Participants: Participants with Fuchs’ endothelial corneal dystrophy. Methods: A validated neural network for image segmentation quantified graft detachment on anterior segment OCT (AS-OCT) images 3 days after DMEK and at the 2-week postoperative visit. Area and volume of graft detachment were compared between DMEK only and triple DMEK using generalized estimating equation models and adjusting for participant age and the size of the air bubble. Main Outcome Measures: Area and volume of DMEK graft detachment. Results: Among 207 participants with 270 eyes included, 75 pseudophakic eyes had DMEK only and 195 eyes had triple DMEK. A total of 147 eyes had less than one third of detachment at day 3. In 139 of these eyes (95%), detachment was still less than one third at the 2-week scan, indicating that postoperative graft detachment at 2 weeks occurred mainly in eyes with early detachment. When superimposing all 3-dimensional maps from 2 weeks after surgery, the central graft was mainly attached and detachment was located at the graft margin. The mean area of graft detachment decreased from 28% in DMEK only and 38% in triple DMEK to 16% in DMEK only and 25% in triple DMEK at the 2-week postoperative visit. At 2 weeks, the mean area of detachment was 1.85-fold higher (95% confidence interval [CI], 1.34–2.56) and the mean volume was 2.41-fold higher (95% CI, 1.51–3.86) in triple DMEK compared with DMEK. A total of 46 eyes received rebubbling procedures, with 7 eyes (9%) in the DMEK group and 39 eyes (20%) in the triple DMEK group (adjusted risk ratio, 3.1; 95% CI, 1.3–7.1), indicating that rebubbling was more common in eyes undergoing triple DMEK. Conclusions: Automated segmentation of AS-OCT images allowed precise quantification of graft detachment over time and identified DMEK combined with cataract surgery as a risk factor. Frequency of operative follow-up might be guided by extent of detachment in the first postoperative days after DMEK.

Purpose: To develop and apply a neural network for quantification of endothelial corneal graft detachment using anterior segment (AS) OCT. Design: Training and validation of a neural network and application within a prospective cohort. Participants: Patients two weeks after Descemet membrane endothelial keratoplasty. Methods: Investigators manually labeled the posterior cornea and the graft in cross-sectional images of rotational AS OCT scans. Neural networks for image segmentation were trained to identify the area of graft detachment on cross-sectional images. The best-performing neural network with the lowest misclassification (Youden index) and highest spatial overlap with the ground truth (Dice coefficient) was selected and evaluated in a separate dataset. Three-dimensional maps of the area and volume of graft detachment were calculated. For application, the neural network’s rating on the detachment was compared with slit-lamp–based ratings of cornea specialists on the same day as the AS OCT imaging took place. Main Outcome Measures: Youden index and Dice coefficient. Results: Neural networks were trained on 27 AS OCT scans with 6912 labeled images. Among 48 combinations of probability thresholds and epoch states, the best-performing neural network showed a Youden index of 0.99 and a Dice coefficient of 0.77, indicating low misclassification and good spatial overlap on individual image segmentation. In the validation set unknown to the neural network with 20 scans (5120 images), the Youden index was 0.85 and the Dice coefficient was 0.73, and a high overall performance compared with the manually labeled ground truth (R2 = 0.90). In the application set with 107 eyes, the neural network estimated the mean percent detachment larger than the cornea specialist (mean difference, 8.2 percentage points; 95% confidence interval, 6.2–10.2). Masked review of 42 AS OCTs with more than ±10 percentage points difference in ratings showed that clinicians underestimated the true detachment in cases with significant detachment requiring intervention. Conclusions: Deep learning-based segmentation of AS OCT images quantified the percent and the volume of DMEK graft detachment with high precision. Fully automated 3-dimensional quantification of graft detachment is highly sensitive, particularly in corneas with a significant amount of detachment, and may support decision making.