Artificial intelligence for eye care Thee, Eric F; Luttikhuizen, Daniël T; Lemij, Hans G ...
Nederlands tijdschrift voor geneeskunde,
09/2020, Volume:
164
Journal Article
Technological developments in ophthalmic imaging and artificial intelligence (AI) create new possibilities for diagnostics in eye care. AI has already been applied in ophthalmic diabetes care. ...AI-systems currently detect diabetic retinopathy in general practice with a high sensitivity and specificity. AI-systems for the screening, monitoring and treatment of age-related macular degeneration and glaucoma are promising and are still being developed. AI-algorithms, however, only perform tasks for which they have been specifically trained and highly depend on the data and reference-standard that were used to train the system in identifying a certain abnormality or disease. How the data and the gold standard were established and determined, influences the performance of the algorithm. Furthermore, interpretability of deep learning algorithms is still an ongoing issue. By highlighting on images the areas that were critical for the decision of the algorithm, users can gain more insight into how algorithms come to a particular result.
The lax eyelid syndrome van den Bosch, W A; Lemij, H G
British journal of ophthalmology,
09/1994, Volume:
78, Issue:
9
Journal Article
Peer reviewed
Open access
The floppy eyelid syndrome (FES) was first described in middle aged, obese men. In later descriptions, age and sex were not specifically mentioned. Associations of FES with various other syndromes ...have been described. The authors question whether all these cases represent the same, single, syndrome. They suggest that a clinical picture similar to FES may occur in lax upper eyelids of any cause. Four such cases are reported here. The authors therefore coin the more general term 'lax eyelid syndrome'. They suggest using the term 'floppy eyelid syndrome' uniquely for patients with the classic signs and symptoms.
Purpose - The purpose of this paper is to explore in a specific hospital care process the applicability in practice of the theories of quality costing and value chains.Design methodology approach - ...In a retrospective case study an in-depth evaluation of the use of a quality cost model (QCM) and the applicability of Porter's care delivery value chain (CDVC) was performed in a specific care process: glaucoma care over the period 2001 to 2006 in the Rotterdam Eye Hospital in The Netherlands.Findings - The case study shows a reduction of costs per product by increasing the number of outpatient visits and surgery combined with a higher patient satisfaction. Reduction of costs of non-compliance by using the QCM is small, due to the absence of (external) financial incentives for both the hospital and individual physicians. For CDVC to be supportive to an integrated quality and cost management the notion "patient value" needs far more specification as mutually agreed on by the stakeholders involved and related reimbursement needs to depend on realised outcomes.Research limitations implications - The case study just focused on one specific care process in one hospital. To determine effects in other areas of health care, it is important to study the use and applicability of the QCM and the CDVC in other care processes and settings.Originality value - QCM and a CDVC can be useful tools for hospital management to manage the outcomes on both quality and costs, but impact is dependent on the incentives in the context of the existing organisational and reimbursement system and asks for an agreed on operationalisation among the various stakeholders of the notion of patient value.
To estimate the corneal endothelial parameters from specular microscopy images depicting cornea guttata (Fuchs dystrophy), we propose a new deep learning methodology that includes a novel attention ...mechanism named feedback non-local attention (fNLA). Our approach first infers the cell edges, then selects the cells that are well detected, and finally applies a postprocessing method to correct mistakes and provide the binary segmentation from which the corneal parameters are estimated (cell density ECD, coefficient of variation CV, and hexagonality HEX). In this study, we analyzed 1203 images acquired with a Topcon SP-1P microscope, 500 of which contained guttae. Manual segmentation was performed in all images. We compared the results of different networks (UNet, ResUNeXt, DenseUNets, UNet++) and found that DenseUNets with fNLA provided the best performance, with a mean absolute error of 23.16 cells/mm\(^{2}\) in ECD, 1.28 % in CV, and 3.13 % in HEX, which was 3-6 times smaller than the error obtained by Topcon's built-in software. Our approach handled the cells affected by guttae remarkably well, detecting cell edges occluded by small guttae while discarding areas covered by large guttae. Overall, the proposed method obtained accurate estimations in extremely challenging specular images.
Saccadic eye movements in Graves' disease Wouters, RJ; van den Bosch, WA; Lemij, HG
Investigative ophthalmology & visual science,
08/1998, Volume:
39, Issue:
9
Journal Article
Peer reviewed
To describe the saccades made by patients with Graves' disease (GD) and to attempt to distinguish these objectively and quantitatively from the saccades in control subjects.
In 12 euthyroid patients ...with GD, the saccades of both eyes were recorded simultaneously with electromagnetic search coils. Subjects were asked to alternate their gazes between two fixed targets that were 10 degrees, 20 degrees, 30 degrees, or 40 degrees apart along each horizontal or vertical meridian. The data from the patients with GD and those from the similarly recorded group of 12 control subjects were examined in two ways. First, the difference in saccadic sizes between the two eyes was assessed. Second, the saccadic dynamics--that is, the maximum velocity and the saturation constant of the main sequence--were determined for each eye. Repeated measurement analysis of variance was used to test observed differences between the two groups. Finally, through exact logistic regression analysis, classification of the saccades as those of a patient with GD or of a control subject was carried out.
The saccades of patients with GD were generally less conjugate than those of control subjects (P < 0.05). On average, the maximum main sequence velocities in patients with GD were lower than in control subjects (P < 0.05). The saccades of patients with GD were well differentiated from those of control subjects.
The saccades in GD may differ markedly from normal saccades, and the two can be reliably distinguished.
The early detection of glaucoma is essential in preventing visual impairment. Artificial intelligence (AI) can be used to analyze color fundus photographs (CFPs) in a cost-effective manner, making ...glaucoma screening more accessible. While AI models for glaucoma screening from CFPs have shown promising results in laboratory settings, their performance decreases significantly in real-world scenarios due to the presence of out-of-distribution and low-quality images. To address this issue, we propose the Artificial Intelligence for Robust Glaucoma Screening (AIROGS) challenge. This challenge includes a large dataset of around 113,000 images from about 60,000 patients and 500 different screening centers, and encourages the development of algorithms that are robust to ungradable and unexpected input data. We evaluated solutions from 14 teams in this paper, and found that the best teams performed similarly to a set of 20 expert ophthalmologists and optometrists. The highest-scoring team achieved an area under the receiver operating characteristic curve of 0.99 (95% CI: 0.98-0.99) for detecting ungradable images on-the-fly. Additionally, many of the algorithms showed robust performance when tested on three other publicly available datasets. These results demonstrate the feasibility of robust AI-enabled glaucoma screening.
Objectives: To evaluate the performance of a deep learning based Artificial Intelligence (AI) software for detection of glaucoma from stereoscopic optic disc photographs, and to compare this ...performance to the performance of a large cohort of ophthalmologists and optometrists. Methods: A retrospective study evaluating the diagnostic performance of an AI software (Pegasus v1.0, Visulytix Ltd., London UK) and comparing it to that of 243 European ophthalmologists and 208 British optometrists, as determined in previous studies, for the detection of glaucomatous optic neuropathy from 94 scanned stereoscopic photographic slides scanned into digital format. Results: Pegasus was able to detect glaucomatous optic neuropathy with an accuracy of 83.4% (95% CI: 77.5-89.2). This is comparable to an average ophthalmologist accuracy of 80.5% (95% CI: 67.2-93.8) and average optometrist accuracy of 80% (95% CI: 67-88) on the same images. In addition, the AI system had an intra-observer agreement (Cohen's Kappa, \(\kappa\)) of 0.74 (95% CI: 0.63-0.85), compared to 0.70 (range: -0.13-1.00; 95% CI: 0.67-0.73) and 0.71 (range: 0.08-1.00) for ophthalmologists and optometrists, respectively. There was no statistically significant difference between the performance of the deep learning system and ophthalmologists or optometrists. There was no statistically significant difference between the performance of the deep learning system and ophthalmologists or optometrists. Conclusion: The AI system obtained a diagnostic performance and repeatability comparable to that of the ophthalmologists and optometrists. We conclude that deep learning based AI systems, such as Pegasus, demonstrate significant promise in the assisted detection of glaucomatous optic neuropathy.
The goal of this article is to investigate the effect of Graves orbitopathy and orbital decompression on the position of the eyeball, and to relate this to a straight upper eyelid contour, observed ...in several patients after correction of upper eyelid retraction. The positions of the pupil center and the upper and lower eyelid margin were measured in relation to a reference line through the medial canthi, on slides obtained from 120 eyelids before and after correction of upper eyelid retraction, and on similar slides obtained from 90 control eyelids. In control subjects, the distance between the reference line and the pupil center was, on average, 4.5 mm (SD 1.2 mm). In decompressed patients, the pupil center was, on average, 2.2 mm lower (SD 1.3 mm), and in nondecompressed patients it was, on average 1.0 mm lower (SD 1.2 mm). A relatively straight upper eyelid contour after otherwise satisfactory correction of upper eyelid retraction was observed in 14 eyelids of 8 patients. All cases occurred in decompressed patients with a distance between the pupil center and the reference line of 2 mm or less. No significant difference was found in the position of the lateral canthal angle between the three groups. Many patients with Graves orbitopathy show a downward shift of the pupil center in relation to a reference line through the medial canthi. This is partially caused by orbital decompression, but also by other factors discussed by the authors. If the distance between this reference line and the pupil center is 2 mm or less, otherwise satisfactory correction of upper eyelid retraction may cause visible straightening of the upper eyelid contour.
PDF
