U-Net Convolutional Neural Network for Real-Time Prediction of the Number of Cultured Corneal Endothelial Cells for Cellular Therapy.

Corneal endothelial decompensation is treated by the corneal transplantation of donor corneas, but donor shortages and other problems associated with corneal transplantation have prompted investigations into tissue engineering therapies. For clinical use, cells used in tissue engineering must undergo strict quality control to ensure their safety and efficacy. In addition, efficient cell manufacturing processes are needed to make cell therapy a sustainable standard procedure with an acceptable economic burden. In this study, we obtained 3098 phase contrast images of cultured human corneal endothelial cells (HCECs). We labeled the images using semi-supervised learning and then trained a model that predicted the cell centers with a precision of 95.1%, a recall of 92.3%, and an F-value of 93.4%. The cell density calculated by the model showed a very strong correlation with the ground truth (Pearson's correlation coefficient = 0.97, p value = 8.10 × 10-52). The total cell numbers calculated by our model based on phase contrast images were close to the numbers calculated using a hemocytometer through passages 1 to 4. Our findings confirm the feasibility of using artificial intelligence-assisted quality control assessments in the field of regenerative medicine.

U-Net Convolutional Neural Network for Segmenting the Corneal Endothelium in a Mouse Model of Fuchs Endothelial Corneal Dystrophy.

PURPOSE: The purpose of this study was to assess the U-Net-based convolutional neural network performance for segmenting corneal endothelium and guttae of Fuchs endothelial corneal dystrophy. METHODS: Twenty-eight images of corneal endothelial cells and guttae of Col8a2L450W/L450W knock-in mice were obtained by specular microscopy. We used 20 images as training data to develop the U-Net for analyzing guttae and cell borders. The proposed network was validated using independent test data of 8 images. Cell density, hexagonality, and coefficient of variation were calculated from the predicted cell borders and compared with ground truth. RESULTS: U-Net allowed the prediction of cell borders and guttae, and overlays of those segmentations on specular microscopy images highly corresponded to ground truth. The average number of guttae per field was 6.25 ± 8.07 for ground truth and 6.25 ± 7.87 when predicted by the network (Pearson correlation coefficient 0.989, P = 3.25 × 10 -6 ). The guttae areas were 1.60% ± 1.79% by manual determination and 1.90% ± 2.02% determined by the network (Pearson correlation coefficient 0.970, P = 6.72 × 10 -5 ). Cell density, hexagonality, and coefficient of variation analyzed by the proposed network for cell borders showed very strong correlations with ground truth (Pearson correlation coefficient 0.989, P = 3.23 × 10 -6 , Pearson correlation coefficient 0.978, P = 2.66 × 10 -5 , and Pearson correlation coefficient 0.936, P = 6.20 × 10 -4 , respectively). CONCLUSIONS: We demonstrated proof of concept for application of U-Net for objective analysis of corneal endothelial cells and guttae in Fuchs endothelial corneal dystrophy, based on limited ground truth data.