Minimally Invasive Corneal Neurotization Provides Sensory Function, Protects Against Recurrent Ulceration, and Improves Visual Acuity.

PURPOSE: To measure sensory recovery after minimally invasive corneal neurotization, and to identify and quantify the extent to which patient and technical factors influence sensory recovery, ulceration rate, and visual outcomes. DESIGN: Retrospective case series. METHODS: This study included 23 patients with neurotrophic keratopathy who underwent indirect corneal neurotization. The primary outcome measure was corneal sensitivity with Cochet-Bonnet aesthesiometry (CBA), and the secondary outcome measure was epithelial breakdown. RESULTS: Over a 7-year period, 28 eyes of 23 patients (mean age, 15.6 ± 13.6 years) were included in the study. The CBA measurements improved from 3.5 ± 9.1 mm at baseline to 44.1 ± 18.2 mm at 24 months after surgery (P < .001). Maximum CBA was reached after 11.1 ± 6.2 months (median, 9 months). Compared to eyes neurotized with a contralateral donor nerve, eyes with an ipsilateral donor nerve achieved a higher mean CBA (36.0 ± 10.9 vs 10.4 ± 14.0 mm, P = .001) at 3 months. Both the number of fascicles (Spearman correlation coefficient, rs -0.474, P = .11) and insertions (rs -0.458, P = .014) negatively correlated with the final CBA. Nine eyes (32.1%) experienced at least 1 episode of epithelial breakdown after surgery. Visual acuity improved in the neurotized corneas from logMAR 0.57 ± 0.79 at baseline to 0.39 ± 0.66 at 12 months (P = .043). CONCLUSIONS: Corneal sensation improves over time after corneal neurotization. There is resultant improvement in visual acuity and protection against epithelial breakdown. It is important to maximize sensory recovery to protect against recurrent ulceration.

Rapid, automated nerve histomorphometry through open-source artificial intelligence.

We aimed to develop and validate a deep learning model for automated segmentation and histomorphometry of myelinated peripheral nerve fibers from light microscopic images. A convolutional neural network integrated in the AxonDeepSeg framework was trained for automated axon/myelin segmentation using a dataset of light-microscopic cross-sectional images of osmium tetroxide-stained rat nerves including various axonal regeneration stages. In a second dataset, accuracy of automated segmentation was determined against manual axon/myelin labels. Automated morphometry results, including axon diameter, myelin sheath thickness and g-ratio were compared against manual straight-line measurements and morphometrics extracted from manual labels with AxonDeepSeg as a reference standard. The neural network achieved high pixel-wise accuracy for nerve fiber segmentations with a mean (± standard deviation) ground truth overlap of 0.93 (± 0.03) for axons and 0.99 (± 0.01) for myelin sheaths, respectively. Nerve fibers were identified with a sensitivity of 0.99 and a precision of 0.97. For each nerve fiber, the myelin thickness, axon diameter, g-ratio, solidity, eccentricity, orientation, and individual x -and y-coordinates were determined automatically. Compared to manual morphometry, automated histomorphometry showed superior agreement with the reference standard while reducing the analysis time to below 2.5% of the time needed for manual morphometry. This open-source convolutional neural network provides rapid and accurate morphometry of entire peripheral nerve cross-sections. Given its easy applicability, it could contribute to significant time savings in biomedical research while extracting unprecedented amounts of objective morphologic information from large image datasets.