An innovative strategy for standardized, structured, and interoperable results in ophthalmic examinations

Analysis of Retinal Thickness in Patients With Chronic Diseases Using Standardized Optical Coherence Tomography Data: Database Study Based on the Radiology Common Data Model

Background

The Observational Medical Outcome Partners-Common Data Model (OMOP-CDM) is an international standard for harmonizing electronic medical record (EMR) data. However, since it does not standardize unstructured data, such as medical imaging, using this data in multi-institutional collaborative research becomes challenging. To overcome this limitation, extensions such as the Radiology Common Data Model (R-CDM) have emerged to include and standardize these data types.

Objective

This work aims to demonstrate that by standardizing optical coherence tomography (OCT) data into an R-CDM format, multi-institutional collaborative studies analyzing changes in retinal thickness in patients with long-standing chronic diseases can be performed efficiently.

Methods

We standardized OCT images collected from two tertiary hospitals for research purposes using the R-CDM. As a proof of concept, we conducted a comparative analysis of retinal thickness between patients who have chronic diseases and those who have not. Patients diagnosed or treated for retinal and choroidal diseases, which could affect retinal thickness, were excluded from the analysis. Using the existing OMOP-CDM at each institution, we extracted cohorts of patients with chronic diseases and control groups, performing large-scale 1:2 propensity score matching (PSM). Subsequently, we linked the OMOP-CDM and R-CDM to extract the OCT image data of these cohorts and analyzed central macular thickness (CMT) and retinal nerve fiber layer (RNFL) thickness using a linear mixed model.

Results

OCT data of 261,874 images from Ajou University Medical Center (AUMC) and 475,626 images from Seoul National University Bundang Hospital (SNUBH) were standardized in the R-CDM format. The R-CDM databases established at each institution were linked with the OMOP-CDM database. Following 1:2 PSM, the type 2 diabetes mellitus (T2DM) cohort included 957 patients, and the control cohort had 1603 patients. During the follow-up period, significant reductions in CMT were observed in the T2DM cohorts at AUMC (P=.04) and SNUBH (P=.007), without significant changes in RNFL thickness (AUMC: P=.56; SNUBH: P=.39). Notably, a significant reduction in CMT during the follow-up was observed only at AUMC in the hypertension cohort, compared to the control group (P=.04); no other significant differences in retinal thickness were found in the remaining analyses.

Conclusions

The significance of our study lies in demonstrating the efficiency of multi-institutional collaborative research that simultaneously uses clinical data and medical imaging data by leveraging the OMOP-CDM for standardizing EMR data and the R-CDM for standardizing medical imaging data.

In-depth analysis of research hotspots and emerging trends in AI for retinal diseases over the past decade

Background

The application of Artificial Intelligence (AI) in diagnosing retinal diseases represents a significant advancement in ophthalmological research, with the potential to reshape future practices in the field. This study explores the extensive applications and emerging research frontiers of AI in retinal diseases.

Objective

This study aims to uncover the developments and predict future directions of AI research in retinal disease over the past decade.

Methods

This study analyzes AI utilization in retinal disease research through articles, using citation data sourced from the Web of Science (WOS) Core Collection database, covering the period from January 1, 2014, to December 31, 2023. A combination of WOS analyzer, CiteSpace 6.2 R4, and VOSviewer 1.6.19 was used for a bibliometric analysis focusing on citation frequency, collaborations, and keyword trends from an expert perspective.

Results

A total of 2,861 articles across 93 countries or regions were cataloged, with notable growth in article numbers since 2017. China leads with 926 articles, constituting 32% of the total. The United States has the highest h-index at 66, while England has the most significant network centrality at 0.24. Notably, the University of London is the leading institution with 99 articles and shares the highest h-index (25) with University College London. The National University of Singapore stands out for its central role with a score of 0.16. Research primarily spans ophthalmology and computer science, with "network," "transfer learning," and "convolutional neural networks" being prominent burst keywords from 2021 to 2023.

Conclusion

China leads globally in article counts, while the United States has a significant research impact. The University of London and University College London have made significant contributions to the literature. Diabetic retinopathy is the retinal disease with the highest volume of research. AI applications have focused on developing algorithms for diagnosing retinal diseases and investigating abnormal physiological features of the eye. Future research should pivot toward more advanced diagnostic systems for ophthalmic diseases.

OCTess: AN OPTICAL CHARACTER RECOGNITION ALGORITHM FOR AUTOMATED DATA EXTRACTION OF SPECTRAL DOMAIN OPTICAL COHERENCE TOMOGRAPHY REPORTS

PURPOSE

Manual extraction of spectral domain optical coherence tomography (SD-OCT) reports is time and resource intensive. This study aimed to develop an optical character recognition (OCR) algorithm for automated data extraction from Cirrus SD-OCT macular cube reports.

METHODS

SD-OCT monocular macular cube reports (n = 675) were randomly selected from a single-center database of patients from 2020 to 2023. Image processing and bounding box operations were performed, and Tesseract (an OCR library) was used to develop the algorithm, OCTess. The algorithm was validated using a separate test data set.

RESULTS

The long short-term memory deep learning version of Tesseract achieved the best performance. After reverifying all discrepancies between human and algorithmic data extractions, OCTess achieved accuracies of 100.00% and 99.98% in the training (n = 125) and testing (n = 550) datasets, while the human error rate was 1.11% (98.89% accuracy) and 0.49% (99.51% accuracy) in each, respectively. OCTess extracted data in 3.1 seconds, compared with 94.3 seconds per report for human evaluators.

CONCLUSION

We developed an OCR and machine learning algorithm that extracted SD-OCT data with near-perfect accuracy, outperforming humans in both accuracy and efficiency. This algorithm can be used for efficient construction of large-scale SD-OCT data sets for researchers and clinicians.

Agreement and precision of wide and cube scan measurements between swept-source and spectral-domain OCT in normal and glaucoma eyes

This study aimed to evaluate agreement of Wide scan measurements from swept-source optical coherence tomography (SS-OCT) Triton and spectral-domain OCT (SD-OCT) Maestro in normal/glaucoma eyes, and to assess the precision of measurements from Wide and Cube scans of both devices. Three Triton and three Maestro operator/device configurations were created by pairing three operators, with study eye and testing order randomized. Three scans were captured for Wide (12 mm × 9 mm), Macular Cube (7 mm × 7 mm-Triton; 6 mm × 6 mm-Maestro), and Optic Disc Cube (6 mm × 6 mm) scans for 25 normal eyes and 25 glaucoma eyes. Parameter measurements included circumpapillary retinal nerve fiber layer(cpRNFL), ganglion cell layer + inner plexiform layer (GCL+), and ganglion cell complex (GCL++). A two-way random effect analysis of variance model was used to estimate the repeatability and reproducibility; agreement was evaluated by Bland-Altman analysis and Deming regression. The precision estimates were low, indicating high precision, for all thickness measurements with the majority of the limits < 5 µm for the macula and < 10 µm for the optic disc. Precision of the Wide and Cube scans were comparable. Excellent agreement between the two devices was found for Wide scans, with the mean difference < 3 µm across all measurements (cpRNFL < 3 µm, GCL+  < 2 µm, GCL ++  < 1 µm), indicating interoperability. A single Wide scan covering the peripapillary and macular regions may be useful for glaucoma diagnosis and management.

Agreement and Precision of Wide and Cube Scan Measurements between Swept-source and Spectral-domain OCT in Normal and Glaucoma Eyes

This study aimed to evaluate agreement of Wide scan measurements from swept-source optical coherence tomography(SS-OCT) Triton and spectral-domain OCT(SD-OCT) Maestro in normal/glaucoma eyes, and to assess the precision of measurements from Wide and Cube scans of both devices. Three Triton and three Maestro operator/device configurations were created by pairing three operators, with study eye and testing order randomized. Three scans were captured for Wide (12mm×9mm), Macular Cube (7mmx7mm-Triton; 6mmx6mm-Maestro), and Optic Disc Cube (6mmx6mm) scans for 25 normal eyes and 25 glaucoma eyes. Thickness of circumpapillary retinal nerve fiber layer(cpRNFL), ganglion cell layer+inner plexiform layer(GCL+), and ganglion cell complex(GCL++) was obtained from each scan. A two-way random effect analysis of variance model was used to estimate the repeatability and reproducibility; agreement was evaluated by Bland-Altman analysis and Deming regression. Precision limit estimates were low: <5µm for macular and <10µm for optic disc parameters. Precision for Wide and Cube scans of both devices were comparablein both groups. Excellent agreement between the two devices was found for Wide scans, with the mean difference<3µm across all measurements (cpRNFL<3µm, GCL+<2µm, GCL++<1µm), indicating interoperability. A single Wide scan covering the peripapillary and macular regions may be useful for glaucoma management.

Towards effective data sharing in ophthalmology: data standardization and data privacy

PURPOSE OF REVIEW

The purpose of this review is to provide an overview of updates in data standardization and data privacy in ophthalmology. These topics represent two key aspects of medical information sharing and are important knowledge areas given trends in data-driven healthcare.

RECENT FINDINGS

Standardization and privacy can be seen as complementary aspects that pertain to data sharing. Standardization promotes the ease and efficacy through which data is shared. Privacy considerations ensure that data sharing is appropriate and sufficiently controlled. There is active development in both areas, including government regulations and common data models to advance standardization, and application of technologies such as blockchain and synthetic data to help tackle privacy issues. These advancements have seen use in ophthalmology, but there are areas where further work is required.

SUMMARY

Information sharing is fundamental to both research and care delivery, and standardization/privacy are key constituent considerations. Therefore, widespread engagement with, and development of, data standardization and privacy ecosystems stand to offer great benefit to ophthalmology.

Real-world treatment intensities and pathways of macular edema following retinal vein occlusion in Korea from Common Data Model in ophthalmology

Despite many studies, optimal treatment sequences or intervals are still questionable in retinal vein occlusion (RVO) macular edema. The aim of this study was to examine the real-world treatment patterns of RVO macular edema. A retrospective analysis of the Observational Medical Outcomes Partnership Common Data Model, a distributed research network, of four large tertiary referral centers (n = 9,202,032) identified 3286 eligible. We visualized treatment pathways (prescription volume and treatment sequence) with sunburst and Sankey diagrams. We calculated the average number of intravitreal injections per patient in the first and second years to evaluate the treatment intensities. Bevacizumab was the most popular first-line drug (80.9%), followed by triamcinolone (15.1%) and dexamethasone (2.28%). Triamcinolone was the most popular drug (8.88%), followed by dexamethasone (6.08%) in patients who began treatment with anti-vascular endothelial growth factor (VEGF) agents. The average number of all intravitreal injections per person decreased in the second year compared with the first year. The average number of injections per person in the first year increased throughout the study. Bevacizumab was the most popular first-line drug and steroids were considered the most common as second-line drugs in patients first treated with anti-VEGF agents. Intensive treatment patterns may cause an increase in intravitreal injections.

Semi-AI and Full-AI digitizer: The ways to digitalize visual field big data

BACKGROUND AND OBJECTIVE

Glaucoma is one of the major diseases that cause blindness, which is incurable and irreversible, and it is essential to detect glaucoma vision deficits in treatment and check the progression of vision disorders in advance. In order to minimize the risk of glaucoma, it is necessary not only to diagnose and observe glaucoma but also to predict prognosis via indicators from Visual Field (VF) tests. However, information from the VF test cannot be directly used in clinical studies because most medical institutions store VF test sheets in Portable Document Format (PDF) or image files in different standards.

METHODS

We developed AI-based real-time VF big data digitizing systems that digitalize VF test images in real-time in two ways; Semi-AI and Full-AI digitizer. The Semi-AI digitizer detects the VF text area with actual coordinates derived from mouse handler system. Full-AI digitizer detects the VF text area with Faster Region Based Convolutional Neural Networks (RCNN). After detecting the text area, both systems extract texts with Recurrent Neural Network based Optical Character Recognition. Semi-AI and Full-AI digitizer post-processes the extracted text results with in-system algorithm and out-of-system algorithm, respectively.

RESULTS

Both systems used 325,310 VF test sheets from a tertiary hospital and extracted a total of 5,530,270 texts. From the 100 randomly selected VF sheets, 3,400 texts were used for the validation. Semi-AI and Full-AI digitizer showed 0.993 and 0.983 of accuracy, respectively.

CONCLUSION

This study demonstrates the effectiveness of AI applications in detecting text areas and the different implementation methodologies of the post-processing process. In detecting text area, Semi-AI may be better than Full-AI digitizer in terms of system speed and human labor labeling if the number of types to be classified is small. However, Full-AI digitizer is recommended because it allows detecting text area regardless of resolution and size of the VF sheets, as the types of real-world VF test sheets cannot be predicted, and the types become more unpredictable when extended to multi-hospital studies. For Post-preprocessing, Semi-AI methodology is recommended because Semi-AI produced higher results with less effort and considered the convenience of researchers by implementing them as in-system.