With recent advancements in machine learning, especially in deep learning, the prediction of eye diseases based on fundus photography using deep convolutional neural networks (DCNNs) has attracted great attention. However, studies focusing on identifying the right disease among several candidates, which is a better approximation of clinical diagnosis in practice comparing with the case that aims to distinguish one particular eye disease from normal controls, are limited. The performance of existing algorithms for multi-class classification of fundus images is at most mediocre. Moreover, in many studies consisting of different eye diseases, labeled images are quite limited mainly due to privacy concern of patients. In this case, it is infeasible to train huge DCNNs, which usually have millions of parameters. To address these challenges, we propose to utilize a lightweight deep learning architecture called MobileNetV2 and transfer learning to distinguish four common eye diseases, including Glaucoma, Maculopathy, Pathological Myopia, and Retinitis Pigmentosa, from normal controls using a small training data. We also apply a visualization approach to highlight the loci that are most related to the disease labels to make the model more explainable. The highlighted area chosen by the algorithm itself may give some hints for further fundus image studies. Our experimental results show that our system achieves an average accuracy of 96.2%, sensitivity of 90.4%, and specificity of 97.6% on the test data via five independent runs, and outperforms two other deep learning-based algorithms both in terms of accuracy and efficiency.

The photoreceptor cell death associated with the various genetic forms of retinitis pigmentosa (RP) is currently untreatable and leads to partial or complete vision loss. Carnosic acid (CA) upregulates endogenous antioxidant enzymes and has proven neuroprotective in studies of neurodegenerative models affecting the brain. In this study, we examined the potential effect of CA on photoreceptor death in the Pde6b^rd10 mouse model of RP. Our data shows that CA provided morphological and functional preservation of photoreceptors. CA appears to exert its neuroprotective effects through inhibition of oxidative stress and endoplasmic reticulum stress.

Photoreceptors consume glucose supplied by the choriocapillaris to support phototransduction and outer segment (OS) renewal. Reduced glucose supply underlies photoreceptor cell death in inherited retinal degeneration and age‐related retinal disease. We have previously shown that restricting glucose transport into the outer retina by conditional deletion of Slc2a1 encoding GLUT1 resulted in photoreceptor loss and impaired OS renewal. However, retinal neurons, glia, and the retinal pigment epithelium play specialized, synergistic roles in metabolite supply and exchange, and the cell‐specific map of glucose uptake and utilization in the retina is incomplete. In these studies, we conditionally deleted Slc2a1 in a pan‐retinal or rod‐specific manner to better understand how glucose is utilized in the retina. Using non‐invasive ocular imaging, electroretinography, and histochemical and biochemical analyses we show that genetic deletion of Slc2a1 from retinal neurons and Müller glia results in reduced OS growth and progressive rod but not cone photoreceptor cell death. Rhodopsin levels were severely decreased even at postnatal day 20 when OS length was relatively normal. Arrestin levels were not changed suggesting that glucose uptake is required to synthesize membrane glycoproteins. Rod‐specific deletion of Slc2a1 resulted in similar changes in OS length and rod photoreceptor cell death. These studies demonstrate that glucose is an essential carbon source for rod photoreceptor cell OS maintenance and viability.

Oxidative stress (OS) and endoplasmic reticulum stress (ERS) are the major factors underlying photoreceptor degeneration. Lutein, RR-zeaxanthin (3R,3’R-zeaxanthin) and RS (meso)-zeaxanthin (3R,3’S-RS- zeaxanthin) (L/Zi) could protect against cell damage by ameliorating OS in retina. In this study, we examined the effect of L/Zi supplementation in a mouse model of photoreceptor degeneration and investigated whether the treatment of L/Zi ameliorated OS and ERS. BALB/cJ mice after light exposure were used as the animal model. The protective effects of L/Zi were observed by electroretinography (ERG) and terminal deoxyuridine triphosphate nick-end labeling (TUNEL) analysis. The underlying mechanisms related to OS and ERS were explored by Western blotting. After L/Zi treatment, the ERG amplitudes were significantly higher, and the number of TUNEL-positive cells was significantly reduced compared to that of the vehicle group. Western blotting results revealed that OS was ameliorated according to the significant downregulation of phosphorylated c-Jun N-terminal kinase (p-JNK), and significant upregulation of nuclear factor erythroid 2-related factor 2 (Nrf2). In addition, ERS was reduced according to the significant downregulation of 78 kDa glucose-regulated protein (GRP78), phosphorylated protein kinase RNA-like endoplasmic reticulum kinase (p-PERK), activating transcription factor 4 (ATF4) and activating transcription factor (ATF6). Our data shows that L/Zi provided functional and morphological preservation of photoreceptors against light damage, which is probably related to its mitigation of oxidative and endoplasmic reticulum stress.

With recent advancements in artificial intelligence, fundus diseases can be classified automatically for early diagnosis, and this is an interest of many researchers. The study aims to detect the edges of the optic cup and the optic disc of fundus images taken from glaucoma patients, which has further applications in the analysis of the cup-to-disc ratio (CDR). We apply a modified U-Net model architecture on various fundus datasets and use segmentation metrics to evaluate the model. We apply edge detection and dilation to post-process the segmentation and better visualize the optic cup and optic disc. Our model results are based on ORIGA, RIM-ONE v3, REFUGE, and Drishti-GS datasets. Our results show that our methodology obtains promising segmentation efficiency for CDR analysis.

Purpose: Retinitis pigmentosa (RP) is a group of inherited diseases characterized by the death of rod photoreceptors, followed by the death of cone photoreceptors, progressively leading to partial or complete blindness. Currently no specific treatment is available for RP patients. Sulforaphane (SFN) has been confirmed to be an effective antioxidant in the treatment of many diseases. In this study, we tested the therapeutic effects of SFN against photoreceptor degeneration in Pde6b^rd10 mice. Methods: rd10 mice and C57/BL6 wild-type (WT) mice were treated with SFN and saline, respectively, from P6 to P20. Electroretinography (ERG), terminal deoxynucleotidyl transferase dUTP nick end labeling and western blot were tested, respectively, at P21 for the analysis of retinal function, retinal cell apoptosis or death and the protein express of GRP78/BiP (TUNEL) as a marker of endoplasmic reticulum (ER) stress. Results: Compared with the saline group, the SFN-treated group showed significantly higher ERG a-wave and b-wave amplitudes, less photoreceptor death, and the downregulation of GRP78/BiP. Conclusions: Our data showed that SFN ameliorated the retinal degeneration of rd10 mice, which is possibly related to the downregulation of GRP78 expression.