We developed an unsupervised machine learning algorithm and applied it to big corneal parameters to identify and monitor keratoconus stages. A big dataset of corneal swept source optical coherence tomography (OCT) images of 12,242 eyes acquired from SS-1000 CASIA OCT Imaging Systems in multiple centers across Japan was assembled. A total of 3,156 eyes with valid Ectasia Status Index (ESI) between zero and 100% were selected for the downstream analysis. Four hundred and twenty corneal topography, elevation, and pachymetry parameters (excluding ESI Keratoconus indices) were selected. The algorithm included three major steps. 1) Principal component analysis (PCA) was used to linearly reduce the dimensionality of the input data from 420 to eight significant principal components. 2) Manifold learning was used to further reducing the selected principal components nonlinearly to two eigen-parameters. 3) Finally, a density-based clustering was applied to the eigen-parameters to identify eyes with keratoconus. Visualization of clusters in 2-D space was used to validate the quality of learning subjectively and ESI was used to assess the accuracy of the identified clusters objectively. The proposed method identified four clusters; I: a cluster composed of mostly normal eyes (224 eyes with ESI equal to zero, 23 eyes with ESI between five and 29, and nine eyes with ESI greater than 29), II: a cluster composed of mostly healthy eyes and eyes with forme fruste keratoconus (1772 eyes with ESI equal to zero, 698 eyes with ESI between five and 29, and 117 eyes with ESI greater than 29), III: a cluster composed of mostly eyes with mild keratoconus stage (184 eyes with ESI greater than 29, 74 eyes with ESI between five and 29, and 6 eyes with ESI equal to zero), and IV: a cluster composed of eyes with mostly advanced keratoconus stage (80 eyes had ESI greater than 29 and 1 eye had ESI between five and 29). We found that keratoconus status and severity can be well identified using unsupervised machine learning algorithms along with linear and non-linear corneal data transformation. The proposed method can better identify and visualize the keratoconus stages.

Supplemental Digital Content is Available in the Text.

No abstract

Aims/hypothesis. Early stage leukocyte entrapment in the retinal microcirculation (retinal leukostasis) is considered to be one of the important pathogenetic events in diabetic retinopathy. Gliclazide, a sulphonylurea, was reported to reduce leukocyte adhesion to endothelial cells in hyperglycaemia in vitro, thus suggesting possible selective efficacy of this sulphonylurea in preventing leukostasis in diabetic patients. This study evaluated the effectiveness and selectivity of gliclazide treatment on retinal leukostasis of diabetic rats in vivo. Methods. Streptozotocin (STZ) (65 mg/kg)-induced diabetic rats were divided into three groups (n = 8 each): an untreated diabetic group, a gliclazide-treated diabetic group, and a glibenclamide-treated diabetic group. Gliclazide or glibenclamide was administered orally during a 3-week period. Non-diabetic rats were used as a control (n = 8). Retinal leukostasis was quantitatively evaluated in vivo by acridine orange leukocyte fluorography using a scanning laser ophthalmoscope.Results. The number of leukocytes trapped in the area around the optic disc in the untreated diabetic group (36.9 ± 5.1 cells) increased significantly compared with the non-diabetic control group (21.9 ± 2.9 cells; p = 0.0007). The number of leukocytes trapped in the gliclazide-treated diabetic group (23.5 ± 4.0 cells) decreased significantly compared with untreated diabetic group (p = 0.0008). In contrast, no reduction of retinal leukostasis was found in the glibenclamide-treated diabetic group (37.8 ± 5.8 cells; p = 0.7923). Conclusion/interpretation. This suggests that gliclazide could directly improve abnormalities in the retinal microcirculation independent of blood glucose control and possibly have selective therapeutic benefits in preventing early, critical events in diabetic retinopathy compared with other sulphonylurea drugs. [Diabetologia (2002) 45:735-739] Keywords Gliclazide, sulphonylurea, diabetic retinopathy, retinal leukostasis, acridine orange leukocyte fluorography, scanning laser ophthalmoscope. Diabetic retinopathy is a leading cause of adult vision loss and blindness and early-stage, preventive strategies are important research areas. Leukocyte adhesion to the diabetic retinal vasculature is presumed to be the critical early event in the pathogenesis of diabetic retinopathy [1,2,3], resulting in a breakdown in the blood-retinal barrier and in capillary nonperfusion [1,4,5,6,7,8,9,10,11,12]. In a rat model of diabetic retinopathy, investigators demonstrated retinal capillary occlusion by neutrophils and monocytes in histologic sections and found that areas of endothelial cell damage, capillary loss, and leukocyte extravasation existed adjacent to the static leukocytes [1]. In another post mortem study of human subjects, increased numbers of neutrophils were observed in the choroid and retina of patients with diabetes [2].

No abstract