Accurate endothelial cell density with specular microscopy is essential for correct clinical assessment of the cornea. Commercial specular microscopes incorporate automated cell segmentation methods to estimate cell density. However, these methods are prone to false cell detections in pathological corneas. This project aims to obtain a reliable automated cell density from specular microscopy images of both healthy and pathological corneas with convolutional neural networks. Convolutional neural networks require labeled datasets. Thus, we developed custom software for producing a curated dataset of labeled ground-truth images and cell density maps. In this paper, we implemented a fully convolutional regression network to predict the cell density map from the input microscopy image. Encouraging preliminary results show the potential of the method. This approach may pave the way for dealing with the variability of corneal endothelial cell images.
Specular microscopy assessment of the human corneal endothelium (CE) in Fuchs’ dystrophy is challenging due to the presence of dark image regions called guttae. This paper proposes a UNet-based segmentation approach that requires minimal post-processing and achieves reliable CE morphometric assessment and guttae identification across all degrees of Fuchs’ dystrophy. We cast the segmentation problem as a regression task of the cell and gutta signed distance maps instead of a pixel-level classification task as typically done with UNets. Compared to the conventional UNet classification approach, the distance-map regression approach converges faster in clinically relevant parameters. It also produces morphometric parameters that agree with the manually-segmented ground-truth data, namely the average cell density difference of -41.9 cells/mm2 (95% confidence interval (CI) [-306.2, 222.5]) and the average difference of mean cell area of 14.8 µm2 (95% CI [-41.9, 71.5]). These results suggest a promising alternative for CE assessment.
We generate synthetic images with a generative adversarial network (GAN) model trained with image patches from specular microscopy corneal endothelial cells. Preliminary results show it may be a suitable approach for reliable cell segmentation.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.