Search

Background: Detecting and classifying the age-related macular degeneration (AMD) at the elderly people is one of the main concerns of the national public health program in Thailand. Within this research purview, accurate and sensitive performance is highly desirable, and a variety of proposed models have been developed toward this end. Deep learning neural networks have recently been shown to have advantage over existing approaches for the task of classifying the level of eye disease. Despite of these advances, there is still significant potential for development, regarding model classification accuracy, and sensitive values. Results: Six very deep learning neural networks (DLNN), named as InceptionV3, ResNet152V2, DenseNet201, EfficientNetB7, InceptionResNetV2, and NASNetLarge are proposed for training to detect and classify the AMD disease from fundus images. The training process with AMD images is implemented based on the transfer learning technique through Google Colab Pro platform. The experiments showed that, for 3-classes AMD classification (Normal, Dry AMD, and Wet AMD), the overall classification accuracy of DenseNet201 is highest and is about 88% with the testing dataset. Furthermore, for the inferred 2-class AMD classification (Normal vs. AMD), the most accuracy values obtained from EfficientNetB7 and InceptionResNetV2 about 93.5% and 95.06%, respectively. Conclusions: The experimental results have shown that, after applying transfer learning technique, the DenseNet201 model had higher accuracy performance in 3-class AMD classification of fundus images than other very deep learning neural networks and other state-of-the-art proposed deep learning models in the literature. Furthermore, the trained EfficientNetB7 and InceptionResNetV2 showed the advantages for the 2-class AMD classification over other existing models.

Supharat Jariyakosol,1,2,&ast; Patcharaporn Jaru-ampornpan,1,2,&ast; Anita Manassakorn,1,2 Rath Itthipanichpong,1,2 Parima Hirunwiwatkul,1,2 Visanee Tantisevi,1,2 Thanapong Somkijrungroj,1,2 Prin Rojanapongpun1,2 1Department of Ophthalmology, Faculty of Medicine, Chulalongkorn University, Bangkok, Thailand; 2Ophthalmology Department, King Chulalongkorn Memorial Hospital, Thai Red Cross Society, Bangkok, Thailand&ast;These authors contributed equally to this workCorrespondence: Supharat Jariyakosol Email jsupharat@gmail.comPurpose: To assess the diagnostic accuracy of visual field results generated by the newly developed software (CU-VF) and the standard automated perimetry (SAP) for detecting hemianopia.Patients and Methods: Forty-three subjects with hemianopia and 33 controls were tested with the CU-VF software on a personal computer and SAP. Hemianopia was defined as the presence of a hemianopic field respecting the vertical meridian on SAP with the corresponding neuroimaging pathology as evaluated by 2 neuro-ophthalmologists. Results of CU-VF were independently evaluated by 2 neuro-ophthalmologists, 1 general ophthalmologist, and 1 general practitioner in terms of the presence of hemianopia. Sensitivity, specificity, and kappa coefficient for inter-observer reliability were calculated. Satisfaction and ease of use were evaluated with a visual analog-scale questionnaire and analyzed using paired t-test.Results: The sensitivity (95% CI) and specificity (95% CI) of the CU-VF to detect hemianopia was 74.42% (58.53– 85.96) and 93.94% (78.38– 99.94). Kappa coefficient between neuro-ophthalmologists versus general ophthalmologist and general practitioner were 0.71 and 0.84, respectively. The mean (SD) test duration was 2.25 (0.002) minutes for the CU-VF and 5.38 (1.34) minutes for SAP (p < 0.001). Subjects reported significantly higher satisfaction and comfort using the CU-VF software compared to SAP.Conclusion: The CU-VF screening software showed good validity and reliability to detect hemianopia, with shorter test duration and higher subject satisfaction compared to SAP.Keywords: hemianopia, visual field, screening software

Glaucoma is a neurodegenerative disease presents with retinal nerve fiber layer (RNFL) defects. We apply machine learning classifiers on the color information of the RNFL to differentiate between intact RNFL (i-RNFL) and RNFL defect (d-RNFL) on optic disc photographs (DPs). DPs from individuals with and without glaucoma were collected. Then, a semi-circle was automatically marked on the DPs, to label i-RNFL versus d-RNFL. RGB intensities and other color spaces of two profiles were collected. Five-fold cross validation is used to compare classification efficiency of five classifiers. A total of 2,051 profiles from 89, 32 and 15 DPs from patients with glaucoma, glaucoma suspects and control subjects were collected. There were 702 and 175 points of d-RNFL and 940 and 234 of i-RNFL in the training and test sets. In the training set, the 3 best classifiers using RGB intensities were fine Gaussian support vector machine (SVM), medium k-Nearest Neighbor and ensemble RUSBoosted Trees, with accuracies of 81.8%, 79.4% and 79.2%. The performance of the fine Gaussian SVM was similar between RGB and other color spaces. In the test set, the highest sensitivity (71.4%) and specificity (88.5%) were archived using RGB and the combination of RGB and Cb and Cr.

Glaucoma is the second leading cause of blindness worldwide. This paper proposes an automated glaucoma screening method using retinal fundus images via the ensemble technique to fuse the results of different classification networks and the result of each classification network was fed as an input to a simple artificial neural network (ANN) to obtain the final result. Three public datasets, i.e., ORIGA-650, RIM-ONE R3, and DRISHTI-GS were used for training and evaluating the performance of the proposed network. The experimental results showed that the proposed network outperformed other state-of-art glaucoma screening algorithms with AUC of 0.94. Our proposed algorithms showed promising potential as a medical support system for glaucoma screening especially in low resource countries.

SIGNIFICANCE Head-mounted video display systems and image processing as a means of enhancing low vision are ideas that have been around for more than 20 years. Recent developments in virtual and augmented reality technology and software have opened up new research opportunities that will lead to benefits for low vision patients. Since the Visionics low vision enhancement system (LVES), the first head-mounted video display LVES, was engineered 20 years ago, various other devices have come and gone with a recent resurgence of the technology over the past few years. In this article, we discuss the history of the development of LVESs, describe the current state of available technology by outlining existing systems, and explore future innovation and research in this area. Although LVESs have now been around for more than two decades, there is still much that remains to be explored. With the growing popularity and availability of virtual reality and augmented reality technologies, we can now integrate these methods within low vision rehabilitation to conduct more research on customized contrast-enhancement strategies, image motion compensation, image-remapping strategies, and binocular disparity, all while incorporating eye-tracking capabilities. Future research should use this available technology and knowledge to learn more about the visual system in the low vision patient and extract this new information to create prescribable vision enhancement solutions for the visually impaired individual.

Background. Assessment of color disc photograph (C-DP) is affected by image quality, which decreases the ability to detect glaucoma. High-dynamic-range (HDR) imaging provides a greater range of luminosity. Therefore, the objective of this study was to evaluate the capability of ophthalmology residents to detect glaucoma using HDR-concept disc photography (HDR-DP) compared to C-DP. Design. Cross-sectional study. Methods. Twenty subjects were classified by 3 glaucoma specialists as either glaucoma, glaucoma suspect, or control. All C-DPs were converted to HDR-DPs and randomly presented and assessed by 10 first-year ophthalmology residents. Sensitivity and specificity of glaucoma detection were compared. Results. The mean ± SD of averaged retinal nerve fiber layer (RNFL) thickness was 74.0 ± 6.1 μm, 100.2 ± 9.6 μm, and 105.8 ± 17.2 μm for glaucoma, glaucoma suspect, and controls, respectively. The diagnostic sensitivity of HDR-DP was higher than that of C-DP (87% versus 68%, mean difference: 19.0, 95% CI: 4.91 to 33.1; p=0.014). Regarding diagnostic specificity, HDR-DP and C-DP yielded 46% and 75% (mean difference: 29.0, 95% CI: 13.4 to 44.6; p=0.002). Conclusions. HDR-DP statistically increased diagnostic sensitivity but not specificity. HDR-DP may be a screening tool for nonexpert ophthalmologists.

Ayurved Siriraj Chantaleela recipe is a traditional Thai remedy consisting of eight medicinal plants, which is employed for the treatment of fever.To investigate the effects of Ayurved Siriraj Herbal recipe Chantaleela on platelet aggregation.Clinical research; ex vivo with before and after study design.Twelve healthy male and female volunteers participated in the present study. Platelet aggregation test before Chantaleela ingestion was done as a control. After administration of 750 mg Chantaleela (3 x 250 mg tablets) every 8 hours for 3 doses, platelet aggregation was measured 8 hours following the first dose using an aggregometer and microplate reader. Adrenaline (Adr) and adenosine diphosphate (ADP) were used as platelet stimulants. Platelet aggregation was measured again at 32 hours and 8-10 days after the first dose.All of the participants completed the present study without any adverse event. Ayurved Siriraj Chantaleela did not affect platelet aggregation; neither Adr nor ADP were used as platelet agonists in both aggregometer and microplate reader Subgroup analysis revealed no significant change in platelet aggregation after Chantaleela administration according to the control for both male and female groups. The same results were also obtained in other subgroup analysis including hyperaggregation group, hypo-normal aggregation group.From the present study, normal dose of Chantaleela for alleviation of fever does not have an effect on either platelet aggregation or platelet numbers. It may conclude that the present study supports the safety use of Chantaleela for relieving fever as platelet status does not need to be taken into consideration.