Taiwan Journal of Ophthalmology最新文献

Visible light optical coherence tomography (VIS-OCT) has made significant progress in the past decade from in vivo proof-of-concept retinal imaging in preclinical models to human clinical translation. The technical advances of VIS-OCT imaging devices include new light sources, optical fiber components, balanced detection methods, and an array of data processing methods. We summarize the unique features of using VIS-OCT in comparison with near-infrared OCT (NIR-OCT), including ultra-high resolution, retinal microvascular oximetry, and reflectance spectroscopy. The ultra-high resolution is granted by the shorter wavelengths in the visible light range ~500-650 nm, as compared with the conventional OCT wavelengths >800 nm. Detailed sub-bandings in the inner plexiform layer and outer segment of photoreceptors, as well as in the retinal pigment epithelium and Bruch's membrane, are consistently resolved in VIS-OCT. The three-dimensional resolving capacity of VIS-OCT allows better isolation of hemoglobin absorption features, allowing blood oxygen saturation (SO₂) calculation in retinal microvasculature. Oximetry calculations were performed down to the capillary level in humans, albeit through massive averaging, which was unattainable by previous methods. Advancing VIS-OCT technology has a high potential to produce significant clinical impact in ophthalmology in the near future.

Retinal diseases often result in photoreceptor dysfunction and cell death, leading to progressive vision impairment and eventual blindness. Clinical management can benefit from assessing photoreceptor mediated visual function, such as for progression monitoring. Meanwhile, therapeutic development requires accurate and reliable vision end points. The emerging optoretinography (ORG) technologies promise noninvasive, objective, and highly sensitive markers of photoreceptor function. ORG is an umbrella term that covers several imaging modalities that use intrinsic optical signal, i.e., without dyes or labeling agents, to quantify photoreceptor responses to light. Early studies showed stimulus-evoked light scattering and morphological changes in isolated photoreceptor and retina, which provided the experimental foundation for subsequent inquires in vivo. Technology advancements allowed for optically quantifying photoreceptor light responses in live human subjects using two-dimensional (2D) fundus photography and 3D optical coherence tomography (OCT). The integration of adaptive optics (AO) with OCT enabled direct measurements of outer segment length changes of individual human photoreceptors, revealing essential steps of the phototransduction cascades. In an effort to enlarge measurement field of view, ease imaging workflow, and improve accessibility, more recent studies investigated ORG techniques without the need to resolve or track individual cells. Clinical ORG imaging culminated in demonstrating highly sensitive and reliable detection of photoreceptor dysfunction in patients with degenerative retinal diseases. Ongoing development of both AO and non-AO ORG approaches promises to advance our understanding of phototransduction and the visual processing pathway, while establishing a powerful clinical tool for assessing vision.

Purpose: This study aimed to develop a high-resolution full-field optical coherence tomography (HR-FFOCT) system for automated three-dimensional (3D) imaging of cultivated epithelial cell sheets and biopsied tissue used in corneal regenerative therapy.

Materials and methods: A commercial HR-FFOCT system (ApolloVue® S100, AMO, Taiwan), originally for dermatological imaging, was re-engineered for <2 μm resolution imaging of rabbit limbal and oral mucosal tissues for limbal stem cell deficiency treatment. Modifications included a piezoelectric transducer for precise Z-stack acquisition, customized LED illumination for registration, and a specialized platform for culture dishes. The system enabled en face and cross-sectional imaging with 3D reconstruction. Rabbit-derived products for cultivated limbal epithelial transplantation (CLET) and cultivated oral mucosal epithelial transplantation (COMET) were imaged before cultivation. Morphology, stratification, and protein expression were analyzed and validated with immunocytochemistry (ICC).

Results: The optimized system produced high-resolution en face and Z-stack images with accurate alignment, capturing stratified epithelial layers in CLET and oral mucosal tissue. Reconstructed 3D images revealed structural detail consistent with ICC-verified expression of junctional proteins, including occludin and actin. Both two-dimensional and 3D visualization of biopsied oral mucosal tissue was achieved. The system enabled noninvasive monitoring of epithelial sheet architecture and thickness without sectioning.

Conclusion: The reconstructed HR-FFOCT system provides a noninvasive, real-time imaging platform for assessing epithelial cell sheets and biopsied tissue in corneal regenerative therapy. It offers potential for standardizing quality evaluation of cell-based products before transplantation and advancing translational applications in regenerative medicine.

Optical coherence tomography (OCT) has revolutionized the diagnosis and management of macular holes (MHs). Before OCT, physicians relied on slit-lamp biomicroscopy and angiographic findings, which were often insufficient for differentiating subtle pathologies. By enabling cross-sectional visualization of the retina, OCT has confirmed vitreomacular traction as a key factor in MH formation and supports the safety of surgical intervention. Quantitative OCT parameters have allowed the development of prognostic biomarkers, while spectral-domain OCT has further improved assessment by revealing postoperative photoreceptor integrity, with restoration of the ellipsoid zone and external limiting membrane correlating with visual recovery. OCT imaging under intraocular gas has enabled early confirmation of MH closure, allowing OCT-guided face-down positioning protocols that reduce postoperative burden without compromising outcomes. En face OCT and three-dimensional volumetric analysis have uncovered new biomarkers, such as preretinal abnormal tissue and inner retinal fluid volume, which aid in surgical planning and visual prognosis. Structural changes, such as dissociation of the optic nerve fiber layer and epiretinal proliferation (EP), are also better understood through OCT, thereby influencing decisions on internal limiting membrane peeling and EP preservation. OCT has not only revolutionized the diagnosis and classification of MHs but also continues to shape their surgical management, bringing us closer to optimizing visual recovery through personalized and data-driven approaches.

Glaucoma is an optic neuropathy and the leading cause of irreversible blindness worldwide. Imaging of the ganglion cell complex and retinal nerve fiber layer with optical coherence tomography (OCT) is a noninvasive, high-resolution means of diagnosing and quantitatively monitoring glaucoma. In the anterior segment, OCT can also be used to assess the anterior chamber angle and identify angle closure, a risk factor for glaucoma. The interpretation of OCT images for accurate diagnosis requires expert-level knowledge of both the technology and glaucoma. Deep learning (DL) is a subfield of artificial intelligence (AI), which is gaining prominence in health care for its ability to interpret images and approximate clinician judgment. This review summarizes recent research that demonstrates how DL can contribute to the analysis of OCT images in glaucoma. Deep neural networks can assist clinicians in checking the quality of OCT scans, quantifying the thickness of optic nerve tissues, evaluating the anterior chamber angle, diagnosing glaucoma, and detecting the progression of existing glaucoma. As further work expands on the generalizability, equity, and explainability of these DL techniques, AI-driven clinical support tools may become available for glaucoma diagnostics.

Purpose: This diagnosis study evaluated the diagnostic validity of Neuro-ophthalmology Disorder Early Screening (NODES), a telemedicine-based smartphone application that integrates structured anamnesis and simple neuro-ophthalmology examination for early detection of neuro-ophthalmological disorders.

Materials and methods: This study was conducted at Sardjito Eye Center Neuro-ophthalmology subdivision, assessing the diagnostic validity of the NODES mobile application of 16 neuro-ophthalmology diagnoses compared to confirmative diagnoses of neuro-ophthalmologists. A total of 126 participants meeting inclusion criteria underwent two treatments: simple examination using the NODES application, compared to confirmative examination by neuro-ophthalmologists. Standard protocols were integrated into both treatments to evaluate the diagnosis accuracy of the examination.

Results: NODES showcase promising results as a new diagnostic tool for these four disorders; myasthenia gravis (MG) with the area under the curve (AUC) =0.951, multicranial nerve palsy orbital apex (MNP OA) with AUC = 0.900, monocranial nerve palsy N.III (MNP N.III) with AUC = 0.824, and monocranial sixth nerve palsy (MNP N.VI) with AUC = 0.764.

Conclusion: The NODES application can be used as a new diagnostic screening tool for several euro-ophthalmology diagnoses including MG, multicranial nerve palsy orbital apex, monocranial nerve palsy N.III, and monocranial sixth nerve palsy. The high specificity of NODES in other neuro-ophthalmology diagnoses can be quite useful to identify individuals without the condition, reducing the occurrence of false positives and crucial in preventing unnecessary stress, further testing, and potential misdiagnoses for those who do not actually have the condition.

Purpose: The purpose of the study was to determine the proportion of diabetic retinopathy (DR) among individuals with type II diabetes mellitus (T2D) and its associated risk factors.

Materials and methods: This was a retrospective analysis of 988 DR outpatients attending a tertiary eye center, Brunei Darussalam, between January 1, 2021, and December 31, 2021.

Results: Eight hundred and fifty one patients (56.8 ± 10.9 years) with a total of 1702 eyes were included. The relative proportion of nonproliferative DR (NPDR) (n = 985, 59.1%) was higher than proliferative DR (PDR) (n = 681, 40.9%), especially in females (32.8%) and Malays (53.4%). The largest DR age groups were NPDR 60-69 years (22.7%) and 50-59 years, with PDR (13.6%). The mean average risk factors between NPDR and PDR only showed marginal differences. The PDR group received more laser treatment (37.2%) than the NPDR group (15.0%). Using a multivariate logistic regression model, visual acuity (VA), measured by logMAR, was significantly associated with differentiating PDR from NPDR. For each single unit increase in logMAR, the risk of PDR increased by 3.58 times (95% confidence interval: 3.04-4.22).

Conclusion: The relative prevalence of NPDR was higher than PDR; however, variables such as hypertension, hyperlipidemia, duration of diabetes, weight, gender, and ethnicity were not strong determinants to differentiate between the two types. VA was associated with progression to PDR. This study highlights the need for early detection, regular screening, and patient education as preventive measures toward the progression of DR.

Purpose: The purpose of this study was to study clinical characteristics, imaging features, and fate of punctate outer retinal toxoplasmosis (PORT).

Materials and methods: A retrospective, observational, descriptive analysis of PORT lesions presenting as satellite lesions of typical full-thickness necrotizing Toxoplasma retinochoroiditis (TRC) or isolated lesions in immunocompetent cases.

Results: We analyzed 34 eyes of 34 cases (22 males and 12 females). PORT lesions appeared as deep, dull, yellowish-gray outer retinal lesions, either as satellite lesions to a TRC (n = 30) or isolated macular punctate lesions (n = 4). The mean lesion size was 562 μm (50-1000). The recurrence rate was high (23.52%), manifesting as either typical TRC (n = 4) or new satellite PORT lesions (n = 4). Active lesions had hyperreflective outer retinal foci on optical coherence tomography (OCT), hyperautofluorescent in the active phase, and stippled upon resolution. Healed lesions exhibited outer nuclear layer thinning and outer retinal atrophy on OCT and were hypoautofluorescent.

Conclusion: PORT lesions were observed as multiple punctate lesions, either adjacent to TRC in active or healed phases or as isolated macular punctate toxoplasmosis. Autofluorescence, angiography, and OCT assist in differentiating active from healed lesions. The fate of PORT lesions was healing with granularity, fading, or complete resolution. Some cases recurred with new satellite PORT lesions, whereas others progressed to typical retinochoroiditis or CNVM.

Purpose: This study aims to evaluate and compare the performance of generative pretrained transformer (GPT)-4o and GPT-4 in answering Taiwan's National Medical Licensing Examination (NMLE) ophthalmology questions from 2014 to 2023, focusing on both answer accuracy and explanation quality.

Materials and methods: A total of 169 ophthalmology questions from Taiwan's NMLE over the past decade were selected. GPT-4o and GPT-4 were tested on each question, and their performance was measured by correct answers and explanations. The results were categorized by ophthalmologic subspecialty and analyzed using statistical methods to determine significant differences between the two models.

Results: GPT-4o achieved a significantly higher overall correct answer rate (92.9%) compared to GPT-4 (69.2%) across all ophthalmology questions from 2014 to 2023 (P < 0.01). GPT-4o outperformed GPT-4 in most subspecialties, including retina (95.8% vs. 58.3%, P < 0.01), external disease and cornea (96.3% vs. 77.8%, P = 0.04), and neuro-ophthalmology (87.5% vs. 50%, P = 0.02). GPT-4o and GPT-4 performed similarly in glaucoma and uveitis, with no significant differences observed. In terms of explanation quality, GPT-4o provided accurate explanations for 90.7% of the questions, with the highest accuracy in pediatric ophthalmology and strabismus (100%) and the lowest in uveitis (83.3%).

Conclusion: GPT-4o exhibited superior performance in both answering and explaining ophthalmology questions from Taiwan's NMLE compared to GPT-4. These results suggest that GPT-4o may be a more reliable tool for educational and diagnostic purposes in ophthalmology.