To develop and evaluate a deep learning system for differentiating between eyes with and without glaucomatous visual field damage (GVFD) and predicting the severity of GFVD from spectral domain OCT ...(SD OCT) optic nerve head images.
Evaluation of a diagnostic technology.
A total of 9765 visual field (VF) SD OCT pairs collected from 1194 participants with and without GVFD (1909 eyes).
Deep learning models were trained to use SD OCT retinal nerve fiber layer (RNFL) thickness maps, RNFL en face images, and confocal scanning laser ophthalmoscopy (CSLO) images to identify eyes with GVFD and predict quantitative VF mean deviation (MD), pattern standard deviation (PSD), and mean VF sectoral pattern deviation (PD) from SD OCT data.
Deep learning models were compared with mean RNFL thickness for identifying GVFD using area under the curve (AUC), sensitivity, and specificity. For predicting MD, PSD, and mean sectoral PD, models were evaluated using R
and mean absolute error (MAE).
In the independent test dataset, the deep learning models based on RNFL en face images achieved an AUC of 0.88 for identifying eyes with GVFD and 0.82 for detecting mild GVFD significantly (P < 0.001) better than using mean RNFL thickness measurements (AUC = 0.82 and 0.73, respectively). Deep learning models outperformed standard RNFL thickness measurements in predicting all quantitative VF metrics. In predicting MD, deep learning models based on RNFL en face images achieved an R
of 0.70 and MAE of 2.5 decibels (dB) compared with 0.45 and 3.7 dB for RNFL thickness measurements. In predicting mean VF sectoral PD, deep learning models achieved high accuracy in the inferior nasal (R
= 0.60) and superior nasal (R
= 0.67) sectors, moderate accuracy in inferior (R
= 0.26) and superior (R
= 0.35) sectors, and lower accuracy in the central (R
= 0.15) and temporal (R
= 0.12) sectors.
Deep learning models had high accuracy in identifying eyes with GFVD and predicting the severity of functional loss from SD OCT images. Accurately predicting the severity of GFVD from SD OCT imaging can help clinicians more effectively individualize the frequency of VF testing to the individual patient.
The ability of deep learning architectures to identify glaucomatous optic neuropathy (GON) in fundus photographs was evaluated. A large database of fundus photographs (n = 14,822) from a racially and ...ethnically diverse group of individuals (over 33% of African descent) was evaluated by expert reviewers and classified as GON or healthy. Several deep learning architectures and the impact of transfer learning were evaluated. The best performing model achieved an overall area under receiver operating characteristic (AUC) of 0.91 in distinguishing GON eyes from healthy eyes. It also achieved an AUC of 0.97 for identifying GON eyes with moderate-to-severe functional loss and 0.89 for GON eyes with mild functional loss. A sensitivity of 88% at a set 95% specificity was achieved in detecting moderate-to-severe GON. In all cases, transfer improved performance and reduced training time. Model visualizations indicate that these deep learning models relied on, in part, anatomical features in the inferior and superior regions of the optic disc, areas commonly used by clinicians to diagnose GON. The results suggest that deep learning-based assessment of fundus images could be useful in clinical decision support systems and in the automation of large-scale glaucoma detection and screening programs.
Full text
Available for:
IZUM, KILJ, NUK, PILJ, PNG, SAZU, UL, UM, UPUK
To investigate the prevalence of visual field defects in glaucomatous eyes, glaucoma suspects, and ocular hypertensives with 24-2 and 10-2 visual fields.
Prospective, cross-sectional study.
Patients ...with or suspected glaucoma tested with 24-2 and 10-2. Patients were classified into 3 groups on the basis of the presence of glaucomatous optic neuropathy (GON) and 24-2 visual field abnormalities: early glaucoma (GON and abnormal visual field, mean deviation >-6 decibels dB), glaucoma suspects (GON and normal visual field), and ocular hypertensives (normal disc, normal visual field, and intraocular pressure >22 mmHg). For the classification of visual field abnormalities, 24-2 and 10-2 tests performed on the same visit were analyzed.
Comparison of the prevalence of abnormal 24-2 versus 10-2 visual field results based on cluster criteria in each diagnostic group.
A total of 775 eyes (497 patients) were evaluated. A total of 364 eyes had early glaucoma, 303 eyes were glaucoma suspects, and 108 eyes were ocular hypertensives. In the glaucoma group, 16 of the 26 eyes (61.5%) classified as normal based on cluster criteria on 24-2 tests were classified as abnormal on 10-2 visual fields. In eyes with suspected glaucoma, 79 of the 200 eyes (39.5%) classified as normal on the 24-2 test were classified as abnormal on 10-2 visual fields. In ocular hypertensive eyes, 28 of the 79 eyes (35.4%) classified as normal on the 24-2 were classified as abnormal on the 10-2. Patients of African descent were more likely to have an abnormal 10-2 result (67.3 vs. 56.8%, P = 0.009).
Central visual field damage seen on the 10-2 test is often missed with the 24-2 strategy in all groups. This finding has implications for the diagnosis of glaucoma and classification of severity.
The post-illumination pupil response (PIPR), which is driven by the intrinsic response of melanopsin-containing, intrinsically photosensitive retinal ganglion cells, has previously been characterized ...in healthy eyes. The present study examined whether the PIPR is affected in patients with glaucoma compared with healthy subjects.
Sixteen glaucoma patients (mean age, 63.7 years) were tested by presenting a 60°, 10-second light stimulus (13 log quanta/cm(2)/s retinal irradiance) of either 470 nm (blue) or 623 nm (red) to one eye after dilation. The consensual pupil response of the fellow undilated eye was recorded by infrared pupillometry for 50 seconds after light offset. These pupillary responses were compared with those of 19 age-matched controls (mean age, 59 years).
The glaucoma patients displayed a net PIPR (blue PIPR minus red PIPR) that was significantly (t-test, P < 0.001) smaller (0.6 mm, SEM 0.12; P < 0.05) than in age-matched controls (1.3 mm, SEM 0.16; P < 0.001). For the patient population, the magnitude of the net PIPR was inversely correlated with the measured visual field loss (mean deviation) of the tested eye.
This study demonstrates that there is a significant decrease in the ipRGC-mediated PIPR in glaucomatous patients when compared to age-matched controls. As the severity of the glaucomatous neuropathy increases, there is a correlated decrease in the PIPR. Therefore, this test has the potential for use as a clinical tool in evaluating patients with glaucoma.
Lamina cribrosa in glaucoma Downs, J Crawford; Girkin, Christopher A
Current opinion in ophthalmology,
2017-March, 2017-Mar, 2017-03-00, 20170301, Volume:
28, Issue:
2
Journal Article
Open access
PURPOSE OF REVIEWThis article presents, summarizes, and interprets the most recent advances in the study and understanding of the lamina cribrosa in glaucoma, in the context of previous work.
RECENT ...FINDINGSThe lamina is an active living structure that responds to strain by changing morphology at the micro-scale and macro-scale in glaucoma. Changes in lamina cribrosa morphology in glaucoma include posteriorization of the laminar insertion into the sclera, increased cupping or depth of the lamina cribrosa, and the development of focal lamina cribrosa defects. These lamina cribrosa changes are associated with disk hemorrhages and visual field damage, and are detectable with clinical imaging techniques such as optical coherence tomography. Glaucomatous changes in the lamina cribrosa are thought to be driven by cellular processes mediated by focal cyclical mechanical strain. Strain is eye specific and mediated by intraocular pressure, cerebrospinal fluid pressure, scleral and lamina cribrosa morphology, and structural stiffness; deleterious lamina cribrosa strains can occur at all levels of mean intraocular pressure.
SUMMARYLaminar morphology is ever changing in health and disease, and recent studies have identified several promising morphological changes that are indicative of glaucoma susceptibility, onset, and progression.
Emerging evidence suggests that the coronavirus disease 2019 (COVID-19) pandemic is disrupting health behaviors such as medication adherence. The objective of this study was to determine whether ...adherence to ocular hypotensive medication was affected by the pandemic and to identify factors associated with this change.
In this cohort study, we used a controlled interrupted time series design in which the interruption was the declaration of the COVID-19 pandemic in the United States on March 13, 2020. The 300-day monitoring period, which evenly bracketed this declaration, started on October 16, 2019, and ended on August 10, 2020.
Patients with primary open-angle glaucoma enrolled in an ongoing longitudinal National Institutes of Health-funded study initiated before the onset of the pandemic were selected if they were prescribed ocular hypotensive medication and had adherence data spanning the 300-day period.
We applied segmented regression analysis using a “slope change following a lag” impact model to obtain the adherence slopes in the periods before and after the segmentation. We compared the 2 slopes using the Davies test.
The main outcome measure was daily adherence to ocular hypotensive medication, defined as the number of doses taken divided by the number of doses prescribed, expressed in percent. Adherence was measured objectively using Medication Event Monitoring System caps. We assessed the associations between change in adherence and demographic, clinical, and psychosocial factors.
The sample included 79 patients (mean age, 71 years standard deviation, 8 years). Segmented regression identified a breakpoint at day 28 after the declaration of the pandemic. The slope in the period after the breakpoint (–0.04%/day) was significantly different from zero (P < 0.001) and from the slope in the period before the breakpoint (0.006%/day; P < 0.001). Mean adherence in the period before the segmentation breakpoint was significantly worse in Black patients (median, IQR: 80.6%, 36.2%) compared with White patients (median, IQR: 97.2%, 8.7%; chi-square, 15.4; P = 0.0004). A significant positive association was observed between the Connor-Davidson resilience score and the change in slope between the periods before and after the breakpoint (P = 0.002).
Adherence to ocular hypotensive medication worsened during the COVID-19 pandemic and seems to be related to patient resilience. This collateral consequence of the pandemic may translate into vision loss that may manifest beyond its containment.
This study presents a quantification method for the assessment of the optic nerve head (ONH) deformations of the living human eye under acute intraocular pressure (IOP) elevation and change of ...cerebrospinal fluid pressure (CSFP) with body position. One eye from a brain-dead organ donor with open-angle glaucoma was imaged by optical coherence tomography angiography during an acute IOP and CSFP elevation test. Volumetric 3D strain was computed by digital volume correlation. With increase in IOP the shear strain consistently increased in both sitting and supine position (p < 0.001). When CSFP was increased at constant IOP by changing body position, a global reduction in the ONH strain was observed (-0.14% p = 0.0264). Strain in the vasculature was significantly higher than in the structural tissue (+0.90%, p = 0.0002). Retinal nerve fiber layer (RNFL) thickness strongly associated (ρ = -0.847, p = 0.008) with strain in the peripapillary sclera (ppScl) but not in the retina (p = 0.433) and lamina (p = 0.611). These initial results show that: CSFP independently to IOP modulates strain in the human ONH; ppScl strains are greater than strains in lamina and retina; strain in the retinal vasculature was higher than in the structural tissue; In this glaucoma eye, higher ppScl strain associated with lower RNFL thickness.
Full text
Available for:
IZUM, KILJ, NUK, PILJ, PNG, SAZU, UL, UM, UPUK
To compare the rates of retinal nerve fiber layer (RNFL) loss in patients suspected of having glaucoma who developed visual field damage (VFD) with those who did not develop VFD and to determine ...whether the rate of RNFL loss can be used to predict the development of VFD.
Prospective, observational cohort study.
Glaucoma suspects, defined as having glaucomatous optic neuropathy or ocular hypertension (intraocular pressure, >21 mmHg) without repeatable VFD at baseline, from the Diagnostic Innovations in Glaucoma Study and the African Descent and Glaucoma Evaluation Study.
Global and quadrant RNFL thickness (RNFLT) were measured with the Spectralis spectral-domain optical coherence tomography (SD-OCT; Spectralis HRA+OCT Heidelberg Engineering, Heidelberg, Germany). Visual field damage was defined as having 3 consecutive abnormal visual fields. The rate of RNFL loss in eyes developing VFD was compared to eyes not developing VFD using multivariate linear mixed-effects models. A joint longitudinal survival model used the estimated RNFLT slope to predict the risk of developing VFD, while adjusting for potential confounding variables.
The rate of RNFL thinning and the probability of developing VFD.
Four hundred fifty-four eyes of 294 glaucoma suspects were included. The average number of SD-OCT examinations was 4.6 (range, 2-9), with median follow-up of 2.2 years (0.4-4.1 years). Forty eyes (8.8%) developed VFD. The estimated mean rate of global RNFL loss was significantly faster in eyes that developed VFD compared with eyes that did not develop VFD (-2.02 μm/year vs. -0.82 μm/year; P<0.001). The joint longitudinal survival model showed that each 1-μm/year faster rate of global RNFL loss corresponded to a 2.05-times higher risk of developing VFD (hazard ratio, 2.05; 95% confidence interval, 1.14-3.71; P = 0.017).
The rate of global RNFL loss was more than twice as fast in eyes that developed VFD compared with eyes that did not develop VFD. A joint longitudinal survival model showed that a 1-μm/year faster rate of RNFLT loss corresponded to a 2.05-times higher risk of developing VFD. These results suggest that measuring the rate of SD-OCT RNFL loss may be a useful tool to help identify patients who are at a high risk of developing visual field loss.
Eye injuries comprise 10–13% of civilian improvised explosive device (IED) injuries. The bomb blast wave induces a normal and shear forces on the tissues, causing a large acute IOP elevation. This ...study calculated the biomechanical stresses and strains in the eye due to IED explosion via eye-specific fluid-structure interaction (FSI) models.
Blast occurred at 2, 3, and 4 m from the front and side of the victim and the weights of the IED were 1 and 2 kg. The ground was covered with the deformable soil to mimic the realistic IED explosion condition and reflect the blast wave.
The IOP elevation of ∼6,000–48,000 mmHg was observed in the eyes while the highest IOP was occurred with the IED weight and distance of 2 kg and 2 m (front) and the lowest was occurred with the IED weight and distance of 1 kg and 4 m (side). Our findings suggest the importance of the victim location and orientation concerning the blast wave when it comes to ocular injury assessment. IOP elevation of ∼2900 and ∼2700 mmHg were observed in ∼1.6 ms after the blast for the IEDS weight of 2 kg and a victim distance of 2 m in front and side blasts, respectively, in consistence with the literature. Nonetheless, IOPs were considerably higher after ∼1.6 ms due to the merging of the bomb blast wave and its reflection off the ground.
The stresses and strains were highest for the frontal blast. Both side and frontal blasts caused higher stresses and strains at the rectus muscle insertions where the sclera is thinnest and prone to rupture. Blast angle has no considerable role in the resultant IOP. Front blast with a heavier IED resulted a higher stresses and deformations in the eye connective tissues compared to the side blast.
Full text
Available for:
GEOZS, IJS, IMTLJ, KILJ, KISLJ, NLZOH, NUK, OILJ, PNG, SAZU, SBCE, SBJE, UILJ, UL, UM, UPCLJ, UPUK, ZAGLJ, ZRSKP
A sustained pupilloconstriction is often observed after the cessation of a bright visual stimulus. This post-illumination pupil response (PIPR) is produced by the intrinsically photosensitive retinal ...ganglion cells (ipRGCs). The present study was designed to examine the characteristics of the PIPR in a normal population without ocular disease.
Thirty-seven subjects (mean age, 48.6 years) were tested by presenting a 60 degrees, 10-second light stimulus (13 log quanta/cm(2)/s retinal irradiance) and recording pupillary responses for 50 seconds after light cessation. The light stimuli (470 blue and 623 red nm) were presented by an optical system to one eye after dilation, while the consensual pupil response of the fellow, undilated eye was recorded by infrared pupillometry.
A positive PIPR was seen in all subjects tested. The population average of the PIPR for 470-nm light was 1.5 mm (SEM 0.10, P < 0.05) and the net PIPR (blue PIPR minus red PIPR) was 1.4 mm (SEM 0.09, P < 0.0001). The net PIPR correlated positively with baseline pupil diameter (P < 0.05), but not significantly with age, race, or sex (P > 0.05) in the test population.
All normal subjects displayed a significant PIPR for a 10-second, 470-nm light stimulus, but not a 623-nm stimulus, which is consistent with the proposed melanopsin-mediated response. In most normal individuals, the amplitude of the PIPR was substantial. This test has the potential to be used as a tool in evaluating subjects with inner retinal dysfunction or melanopsin-related disorders.
PDF
