Search

Abstract Multiple glaucoma studies have attempted to generate visual field (VF) mean deviation (MD) estimates using cross-sectional optical coherence tomography (OCT) data. However, whether such models offer any value in detecting longitudinal VF progression is unclear. We address this by developing a machine learning (ML) model to convert OCT data to MD and assessing its ability to detect longitudinal worsening. In this study, we created a model dataset of 70,575 paired OCT/VFs to train an ML model to convert OCT to VF-MD. We created a separate progression dataset of 4,044 eyes with ≥ 5 paired OCT/VFs to assess the ability of OCT-derived MD to detect worsening. The progression dataset eyes had 2 additional unpaired VFs (≥ 7 total) to establish a “ground truth” rate of progression defined by MD slope. We used the ML model to generate longitudinal OCT-MD estimates for each OCT scan for progression dataset eyes. We calculated MD slopes after substituting/supplementing VF-MD with OCT-MD and measured the ability to detect progression. We labeled true progressors using a ground truth MD slope

Abstract To develop and evaluate the performance of a deep learning model (DLM) that predicts eyes at high risk of surgical intervention for uncontrolled glaucoma based on multimodal data from an initial ophthalmology visit. Longitudinal, observational, retrospective study. 4898 unique eyes from 4038 adult glaucoma or glaucoma-suspect patients who underwent surgery for uncontrolled glaucoma (trabeculectomy, tube shunt, xen, or diode surgery) between 2013 and 2021, or did not undergo glaucoma surgery but had 3 or more ophthalmology visits. We constructed a DLM to predict the occurrence of glaucoma surgery within various time horizons from a baseline visit. Model inputs included spatially oriented visual field (VF) and optical coherence tomography (OCT) data as well as clinical and demographic features. Separate DLMs with the same architecture were trained to predict the occurrence of surgery within 3 months, within 3–6 months, within 6 months–1 year, within 1–2 years, within 2–3 years, within 3–4 years, and within 4–5 years from the baseline visit. Included eyes were randomly split into 60%, 20%, and 20% for training, validation, and testing. DLM performance was measured using area under the receiver operating characteristic curve (AUC) and precision-recall curve (PRC). Shapley additive explanations (SHAP) were utilized to assess the importance of different features. Model prediction of surgery for uncontrolled glaucoma within 3 months had the best AUC of 0.92 (95% CI 0.88, 0.96). DLMs achieved clinically useful AUC values (> 0.8) for all models that predicted the occurrence of surgery within 3 years. According to SHAP analysis, all 7 models placed intraocular pressure (IOP) within the five most important features in predicting the occurrence of glaucoma surgery. Mean deviation (MD) and average retinal nerve fiber layer (RNFL) thickness were listed among the top 5 most important features by 6 of the 7 models. DLMs can successfully identify eyes requiring surgery for uncontrolled glaucoma within specific time horizons. Predictive performance decreases as the time horizon for forecasting surgery increases. Implementing prediction models in a clinical setting may help identify patients that should be referred to a glaucoma specialist for surgical evaluation.

Linear regression of optical coherence tomography measurements of peripapillary retinal nerve fiber layer thickness is often used to detect glaucoma progression and forecast future disease course. However, current measurement frequencies suggest that clinicians often apply linear regression to a relatively small number of measurements (e.g., less than a handful). In this study, we estimate the accuracy of linear regression in predicting the next reliable measurement of average retinal nerve fiber layer thickness using Zeiss Cirrus optical coherence tomography measurements of average retinal nerve fiber layer thickness from a sample of 6,471 eyes with glaucoma or glaucoma-suspect status. Linear regression is compared to two null models: no glaucoma worsening, and worsening due to aging. Linear regression on the first M ≥ 2 measurements was significantly worse at predicting a reliable M+1st measurement for 2 ≤ M ≤ 6. This range was reduced to 2 ≤ M ≤ 5 when retinal nerve fiber layer thickness measurements were first "corrected" for scan quality. Simulations based on measurement frequencies in our sample-on average 393 ± 190 days between consecutive measurements-show that linear regression outperforms both null models when M ≥ 5 and the goal is to forecast moderate (75th percentile) worsening, and when M ≥ 3 for rapid (90th percentile) worsening. If linear regression is used to assess disease trajectory with a small number of measurements over short time periods (e.g., 1-2 years), as is often the case in clinical practice, the number of optical coherence tomography examinations needs to be increased.

Abstract Glaucoma is a leading cause of irreversible blindness, and its worsening is most often monitored with visual field (VF) testing. Deep learning models (DLM) may help identify VF worsening consistently and reproducibly. In this study, we developed and investigated the performance of a DLM on a large population of glaucoma patients. We included 5099 patients (8705 eyes) seen at one institute from June 1990 to June 2020 that had VF testing as well as clinician assessment of VF worsening. Since there is no gold standard to identify VF worsening, we used a consensus of six commonly used algorithmic methods which include global regressions as well as point-wise change in the VFs. We used the consensus decision as a reference standard to train/test the DLM and evaluate clinician performance. 80%, 10%, and 10% of patients were included in training, validation, and test sets, respectively. Of the 873 eyes in the test set, 309 [60.6%] were from females and the median age was 62.4; (IQR 54.8–68.9). The DLM achieved an AUC of 0.94 (95% CI 0.93–0.99). Even after removing the 6 most recent VFs, providing fewer data points to the model, the DLM successfully identified worsening with an AUC of 0.78 (95% CI 0.72–0.84). Clinician assessment of worsening (based on documentation from the health record at the time of the final VF in each eye) had an AUC of 0.64 (95% CI 0.63–0.66). Both the DLM and clinician performed worse when the initial disease was more severe. This data shows that a DLM trained on a consensus of methods to define worsening successfully identified VF worsening and could help guide clinicians during routine clinical care.

Marine big data are characterized by a large amount and complex structures, which bring great challenges to data management and retrieval. Based on the GeoSOT Grid Code and the composite index structure of the MongoDB database, this paper proposes a spatio-temporal grid index model (STGI) for efficient optimized query of marine big data. A spatio-temporal secondary index is created on the spatial code and time code columns to build a composite index in the MongoDB database used for the storage of massive marine data. Multiple comparative experiments demonstrate that the retrieval efficiency adopting the STGI approach is increased by more than two to three times compared with other index models. Through theoretical analysis and experimental verification, the conclusion could be achieved that the STGI model is quite suitable for retrieving large-scale spatial data with low time frequency, such as marine big data.

As the amount of collected spatial information (2D/3D) increases, the real-time processing of these massive data is among the urgent issues that need to be dealt with. Discretizing the physical earth into a digital gridded earth and assigning an integral computable code to each grid has become an effective way to accelerate real-time processing. Researchers have proposed optimization algorithms for spatial calculations in specific scenarios. However, a complete set of algorithms for real-time processing using grid coding is still lacking. To address this issue, a carefully designed, integral grid-coding algebraic operation framework for GeoSOT-3D (a multilayer latitude and longitude grid model) is proposed. By converting traditional floating-point calculations based on latitude and longitude into binary operations, the complexity of the algorithm is greatly reduced. We then present the detailed algorithms that were designed, including basic operations, vector operations, code conversion operations, spatial operations, metric operations, topological relation operations, and set operations. To verify the feasibility and efficiency of the above algorithms, we developed an experimental platform using C++ language (including major algorithms, and more algorithms may be expanded in the future). Then, we generated random data and conducted experiments. The experimental results show that the computing framework is feasible and can significantly improve the efficiency of spatial processing. The algebraic operation framework is expected to support large geospatial data retrieval and analysis, and experience a revival, on top of parallel and distributed computing, in an era of large geospatial data.

With the rapid development of large-scale satellite constellations and the increasing demand for rapid communication and emergency rescue using global satellite-based Internet, there have been new requirements for efficient algorithms for inter-communication between satellites. As the constellations of low-orbit satellites become larger, the complexities of real-time inter-satellite calculation and path planning are becoming more complicated and are increasing geometrically. To address the bottlenecks in large-scale node space computing, we introduced a global space grid. Based on this grid, an efficient calculation method of spatial inter-connection between satellite constellations is proposed, according to the concept of “storage for computing” and the high computational efficiency of the spatial grid model. This strategy includes the following parts: (1) the introduction of the GeoSOT-3D global grid model into aerospace and the construction of the aerospace grid indexing BigTable; (2) a set of algorithms for satellite visibility analysis according to the visible grid look-up table and the secondary grid index; and (3) planning inter-satellite routing by querying the grid’s inherent visibility. The idea at the basis of this method is to employ the “space for time” concept to convert the high-dimensional floating operations into one-dimensional matching operations by querying the inherent “visible” attribute of the grid. In our study, we simulated thousands of satellites, discretized their trajectories into grids, and pre-calculated the visibility between grid cells to plan the routing path for the ground data transmission. The theoretical analysis and experimental verification show that the algorithm is feasible and efficient, and it can significantly improve the computational efficiency of inter-satellite connection. We hope that the method can be used in emergency communications, disaster warning, and maritime rescue, and can contribute to the next generation of satellite internet and “satellite-ground” integrated networks.

Based on the analysis of the problems in the generation algorithm of discrete grid systems domestically and abroad, a new universal algorithm for the unit duplication of a polyhedral discrete grid is proposed, and its core is “simple unit replication + effective region restriction”. First, the grid coordinate system and the corresponding spatial rectangular coordinate system are established to determine the rectangular coordinates of any grid cell node. Then, the type of the subdivision grid system to be calculated is determined to identify the three key factors affecting the grid types, which are the position of the starting point, the length of the starting edge, and the direction of the starting edge. On this basis, the effective boundary of a multiscale grid can be determined and the grid coordinates of a multiscale grid can be obtained. A one-to-one correspondence between the multiscale grids and subdivision types can be established. Through the appropriate rotation, translation and scaling of the multiscale grid, the node coordinates of a single triangular grid system are calculated, and the relationships between the nodes of different levels are established. Finally, this paper takes a hexagonal grid as an example to carry out the experiment verifications by converting a single triangular grid system (plane) directly to a single triangular grid with a positive icosahedral surface to generate a positive icosahedral surface grid. The experimental results show that the algorithm has good universality and can generate the multiscale grid of an arbitrary grid configuration by adjusting the corresponding starting transformation parameters.