DenseUNets with feedback non-local attention for the segmentation of specular microscopy images of the corneal endothelium with guttae

Abstract

Corneal guttae, which are the abnormal growth of extracellular matrix in the corneal endothelium, are observed in specular images as black droplets that occlude the endothelial cells. To estimate the corneal parameters (endothelial cell density [ECD], coefficient of variation [CV], and hexagonality [HEX]), we propose a new deep learning method that includes a novel attention mechanism (named fNLA), which helps to infer the cell edges in the occluded areas. The approach first derives the cell edges, then infers the well-detected cells, and finally employs a postprocessing method to fix mistakes. This results in a binary segmentation from which the corneal parameters are estimated. We analyzed 1203 images (500 contained guttae) obtained with a Topcon SP-1P microscope. To generate the ground truth, we performed manual segmentation in all images. Several networks were evaluated (UNet, ResUNeXt, DenseUNets, UNet++, etc.) and we found that DenseUNets with fNLA provided the lowest error: a mean absolute error of 23.16 [cells/mm$$^{2}$$ 2 ] in ECD, 1.28 [%] in CV, and 3.13 [%] in HEX. Compared with Topcon’s built-in software, our error was 3–6 times smaller. Overall, our approach handled notably well the cells affected by guttae, detecting cell edges partially occluded by small guttae and discarding large areas covered by extensive guttae.

Corneal Guttae After Descemet Membrane Endothelial Keratoplasty

“…22 Corneal guttae are known to affect endothelial cells, 3 but the effect of the observed corneal guttae on postoperative ECD is difficult to quantify because endothelial cell counts on specular microscopy images in the presence of corneal guttae do not tend to be reliable due to a high degree of variability in cell density. 23,24 For best-corrected visual acuity, 10 of 11 eyes without graft failure had a bestcorrected visual acuity of 0.9 (decimal) or better at the last available follow-up which is comparable to DMEK outcomes reported earlier. 22,25,26 In conclusion, corneal guttae can occur after DMEK including in eyes operated for indications other than FECD.…”

Deep learning segmentation of endothelial cell images using an active learning paradigm with guided label corrections

“…In the application of corneal EC image segmentation, deep learning studies only cover methods to automatically analyze EC images using U-Net and its variations 11 , 13 – 16 A few studies in other applications, such as liver and spleen segmentation in CT images, have reported human-in-the-loop and reinforcement learning approaches as means to create large quantities of ground truth segmentations 26 . Our study builds on the literature by exploring a different deep learning network, DeepLabV3+, and its enhanced segmentation performance on post-DMEK EC images compared to U-Net.…”

“…Most recently, Vigueras-Guillén et al 13 – 16 developed enhanced U-Net models with combined region of interest and edge detection, dense blocks, feature map rotations and reflections, and feedback nonlocal attention blocks.…”

Artificial Intelligence for Anterior Segment Diseases: A Review of Potential Developments and Clinical Applications

“…Vigueras-Guillén et al. developed DL methods to estimate the corneal endothelium parameters using 500 SM images with guttae [ 39 ]. The DL methods obtained lower mean absolute errors compared with commercial software.…”