We aimed to develop a model to predict visual field (VF) in the central 10 degrees in patients with glaucoma, by training a convolutional neural network (CNN) with optical coherence tomography (OCT) ...images and adjusting the values with Humphrey Field Analyzer (HFA) 24-2 test. The training dataset included 558 eyes from 312 glaucoma patients and 90 eyes from 46 normal subjects. The testing dataset included 105 eyes from 72 glaucoma patients. All eyes were analyzed by the HFA 10-2 test and OCT; eyes in the testing dataset were additionally analyzed by the HFA 24-2 test. During CNN model training, the total deviation (TD) values of the HFA 10-2 test point were predicted from the combined OCT-measured macular retinal layers' thicknesses. Then, the predicted TD values were corrected using the TD values of the innermost four points from the HFA 24-2 test. Mean absolute error derived from the CNN models ranged between 9.4 and 9.5 B. These values reduced to 5.5 dB on average, when the data were corrected using the HFA 24-2 test. In conclusion, HFA 10-2 test results can be predicted with a OCT images using a trained CNN model with adjustment using HFA 24-2 test.
Full text
Available for:
IZUM, KILJ, NUK, PILJ, PNG, SAZU, UL, UM, UPUK
To evaluate the effect of elongated photoreceptor outer segment length on the visual prognosis of patients with chronic central serous chorioretinopathy after treatment using half-dose and ...half-fluence photodynamic therapy (reduced PDT).
The study included 36 eyes of 36 patients with chronic central serous chorioretinopathy who underwent reduced PDT and were followed up for at least 1 year. Spectral domain optical coherence tomography measurement was conducted at baseline and 12 months after reduced PDT. Thereafter, the association between the best-corrected visual acuity (BCVA) at 12 months after reduced PDT and 7 baseline variables (age, symptom duration, BCVA, outer nuclear layer thickness, elongated photoreceptor outer segment length, height of subretinal detachment, and subfoveal choroidal thickness) was evaluated. Multivariate analyses using the model selection with the corrected Akaike Information Criterion index were performed.
The optimal model for BCVA at 12 months only included baseline BCVA (coefficient = 0.90, P < 0.0001) and baseline elongated photoreceptor outer segment length (coefficient = -0.0016, P = 0.034), but not outer nuclear layer thickness.
Elongated photoreceptor outer segment length was significantly associated with BCVA prognosis in patients with chronic central serous chorioretinopathy after reduced PDT and can be useful for predicting residual photoreceptor function during the active phase of chronic central serous chorioretinopathy.
To investigate the clinical validity of the Guided Progression Analysis definition (GPAD) and cluster-based definition (CBD) with the Humphrey Field Analyzer (HFA) 10-2 test in retinitis pigmentosa ...(RP). Ten non-progressive RP visual fields (VFs) (HFA 10-2 test) were simulated for each of 10 VFs of 111 eyes (10 simulations x 10 VF sequencies x 111 eyes = 111,000 VFs; Dataset 1). Using these simulated VFs, the specificity of GPAD for the detection of progression was determined. Using this dataset, similar analyses were conducted for the CBD, in which the HFA 10-2 test was divided into four quadrants. Subsequently, the Hybrid Definition was designed by combining the GPAD and CBD; various conditions of the GPAD and CBD were altered to approach a specificity of 95.0%. Subsequently, actual HFA 10-2 tests of 116 RP eyes (10 VFs each) were collected (Dataset 2), and true positive rate, true negative rate, false positive rate, and the time required to detect VF progression were evaluated and compared across the GPAD, CBD, and Hybrid Definition. Specificity values were 95.4% and 98.5% for GPAD and CBD, respectively. There were no significant differences in true positive rate, true negative rate, and false positive rate between the GPAD, CBD, and Hybrid Definition. The GPAD and Hybrid Definition detected progression significantly earlier than the CBD (at 4.5, 5.0, and 4.5 years, respectively). The GPAD and the optimized Hybrid Definition exhibited similar ability for the detection of progression, with the specificity reaching 95.4%.
Full text
Available for:
DOBA, IZUM, KILJ, NUK, PILJ, PNG, SAZU, SIK, UILJ, UKNU, UL, UM, UPUK
We examined the potential association of idiopathic normal pressure hydrocephalus (iNPH) with the generation of normal-tension glaucoma (NTG), to explore possible relationships between intracranial ...pressure (ICP) and the presence of glaucoma, and to compare disc morphology of NTG patients with or without iNPH. We investigated 20 iNPH patients, examined the prevalence of glaucoma, and compared the optic discs of NTG patients with iNPH (n = 11) and age-matched NTG patients without iNPH (n = 16). All data were collected prior to the treatment of iNPH, to eliminate the possibility that the treatment may have contributed to the progression of NTG. The diagnoses of NTG were made using visual field data, intraocular pressure measurements, fundoscopy, and optical coherence tomography (OCT). Using OCT, the optic nerve disc depth was also measured. The ICP was higher in the iNPH with NTG compared to iNPH without NTG (p = 0.0425), and the cupping depths of the discs of NTG patients with iNPH were significantly shallower compared with those of NTG patients without iNPH (p = 0.0097). Based on the difference in cupping depth, NTG patients with iNPH may have a different morphology from typical glaucoma patients, which could in turn reflect a different pathogenesis compared to NTG patients without iNPH.
Full text
Available for:
IZUM, KILJ, NUK, PILJ, PNG, SAZU, UL, UM, UPUK
Pachychoroid neovasculopathy (PNV) shares some anatomical features with other pachychoroid spectrum diseases, but little is known about the characteristics on the treatment with anti-vascular ...endothelial growth factor (VEGF). We investigated the effect of choroidal structure and responses to anti-VEGF on the prognosis of pachychoroid neovasculopathy (PNV) and other types of neovascular age-related macular degeneration (non-PNV). Twenty-one eyes with PNV and 34 eyes with non-PNV who had anti-VEGF treatment were retrospectively reviewed. Choroidal neovascularization (CNV) area at baseline was measured with fluorescein angiography (FAG). The luminal and stromal area in the choroid was measured by enhanced-depth-imaging (EDI) OCT at baseline and 1 month. The association between dry macula or LogMAR VA (visual acuity, VA) at 1 month and baseline values or changes in the luminal or stromal area at 1 month, baseline CNV area, or anti-VEGF drugs were analyzed in patients with or without PNV. In non-PNV, change of luminal area (coefficient = 7.0×10-5, p = 0.0001), baseline CNV area (coefficient = 0.18, p = 0.033), and aflibercept vs. ranibizumab (coefficient = 0.29, p = 0.0048) were chosen as predictors for dry macula by the model selection. Similarly, in non-PNV, change of luminal area (coefficient = 6.1×10-6, p = 0.033), baseline CNV area (coefficient = 0.034, p = 0.022), and aflibercept vs. ranibizumab (coefficient = 0.056, p = 0.0020) were chosen as predictors for greater VA improvement. In PNV, however, none of these factors was chosen as predictors for dry macula or VA improvement by the model selection. The result of the present study implied that structural response after anti-VEGF might be different between non-PNV and PNV in the treatment with anti-VEGF agents.
Full text
Available for:
DOBA, IZUM, KILJ, NUK, PILJ, PNG, SAZU, SIK, UILJ, UKNU, UL, UM, UPUK
To predict the visual field (VF) of glaucoma patients within the central 10° from optical coherence tomography (OCT) measurements using deep learning and tensor regression.
Cross-sectional study.
...Humphrey 10-2 VFs and OCT measurements were carried out in 505 eyes of 304 glaucoma patients and 86 eyes of 43 normal subjects. VF sensitivity at each test point was predicted from OCT-measured thicknesses of macular ganglion cell layer + inner plexiform layer, retinal nerve fiber layer, and outer segment + retinal pigment epithelium. Two convolutional neural network (CNN) models were generated: (1) CNN-PR, which simply connects the output of the CNN to each VF test point; and (2) CNN-TR, which connects the output of the CNN to each VF test point using tensor regression. Prediction performance was assessed using 5-fold cross-validation through the root mean squared error (RMSE). For comparison, RMSE values were also calculated using multiple linear regression (MLR) and support vector regression (SVR). In addition, the absolute prediction error for predicting mean sensitivity in the whole VF was analyzed.
RMSE with the CNN-TR model averaged 6.32 ± 3.76 (mean ± standard deviation) dB. Significantly (P < .05) larger RMSEs were obtained with other models: CNN-PR (6.76 ± 3.86 dB), SVR (7.18 ± 3.87 dB), and MLR (8.56 ± 3.69 dB). The absolute mean prediction error for the whole VF was 2.72 ± 2.60 dB with the CNN-TR model.
The Humphrey 10-2 VF can be predicted from OCT-measured retinal layer thicknesses using deep learning and tensor regression.
Full text
Available for:
GEOZS, IJS, IMTLJ, KILJ, KISLJ, NLZOH, NUK, OILJ, PNG, SAZU, SBCE, SBJE, UILJ, UL, UM, UPCLJ, UPUK, ZAGLJ, ZRSKP
The aim of the study was to investigate the usefulness of processing visual field (VF) using a variational autoencoder (VAE). The training data consisted of 82,433 VFs from 16,836 eyes. Testing ...dataset 1 consisted of test-retest VFs from 104 eyes with open angle glaucoma. Testing dataset 2 was series of 10 VFs from 638 eyes with open angle glaucoma. A VAE model to reconstruct VF was developed using the training dataset. VFs in the testing dataset 1 were then reconstructed using the trained VAE and the mean total deviation (mTD) was calculated (mTD
). In testing dataset 2, the mTD value of the tenth VF was predicted using shorter series of VFs. A similar calculation was carried out using a weighted linear regression where the weights were equal to the absolute difference between mTD and mTD
. In testing dataset 1, there was a significant relationship between the difference between mTD and mTD
from the first VF and the difference between mTD in the first and second VFs. In testing dataset 2, mean squared prediction errors with the weighted mTD trend analysis were significantly smaller than those form the unweighted mTD trend analysis.
Full text
Available for:
IZUM, KILJ, NUK, PILJ, PNG, SAZU, UL, UM, UPUK
Short wavelength automated perimetry (SWAP) is known for detecting the early reduction of retinal sensitivity (RS) in glaucoma. It's application in retinal diseases have also been discussed ...previously. We investigated the difference in RS measured between standard white-on-white automated perimetry (WW) and blue-on-yellow SWAP in central serous chorioretinopathy (CSC). The overall RS (W-RS, S-RS) as well as the RS inside and outside of the serous retinal detachment (SRD) region were investigated in 26 eyes of 26 CSC patients using WW and SWAP. The central retinal thickness, central choroidal thickness, SRD area (SRDa), and SRD height at the fovea were measured using optic coherence tomography. RS inside the SRD region was lower than that of outside for both perimetries (both p < 0.001). The difference between RS inside and outside of the SRD region was greater in SWAP compared to WW (p < 0.001). Univariate analysis revealed significant correlations between SRDa and both W-RS and S-RS (both p < 0.001); moreover, multivariate analysis indicated that only S-RS was selected as the optimal model for SRDa. Our study demonstrated that SWAP was detected the decrease in RS more accurately than WW in CSC. These results may suggest the usefulness of SWAP for detecting change of retinal function in CSC.
Full text
Available for:
IZUM, KILJ, NUK, PILJ, PNG, SAZU, UL, UM, UPUK
The aim of the current study is to identify possible new Ocular Response Analyzer (ORA) waveform parameters related to changes of retinal structure/deformation, as measured by the peripapillary ...retinal arteries angle (PRAA), using a generative deep learning method of variational autoencoder (VAE). Fifty-four eyes of 52 subjects were enrolled. The PRAA was calculated from fundus photographs and was used to train a VAE model. By analyzing the ORA waveform reconstructed (noise filtered) using VAE, a novel ORA waveform parameter (Monot1-2), was introduced, representing the change in monotonicity between the first and second applanation peak of the waveform. The variables mostly related to the PRAA were identified from a set of 41 variables including age, axial length (AL), keratometry, ORA corneal hysteresis, ORA corneal resistant factor, 35 well established ORA waveform parameters, and Monot1-2, using a model selection method based on the second-order bias-corrected Akaike information criterion. The optimal model for PRAA was the AL and six ORA waveform parameters, including Monot1-2. This optimal model was significantly better than the model without Monot1-2 (p = 0.0031, ANOVA). The current study suggested the value of a generative deep learning approach in discovering new useful parameters that may have clinical relevance.
Full text
Available for:
IZUM, KILJ, NUK, PILJ, PNG, SAZU, UL, UM, UPUK
To investigate retinal vessel shift (RVS) and its association with axial length (AL) elongation in junior high school students.
Total 161 eyes of 161 healthy junior high school students were ...prospectively studied. Optical AL and anterior chamber depth (ACD) measurements, and fundus photography were performed in the first and third grades. Eyes of subjects in the first and third grade that had perfect matching among all the retinal vessels were allocated to the RVS(-) group, otherwise allocated to the RVS(+) group. In the RVS(+) group, the peripapillary retinal arteries angle (PRAA) was measured for quantitative analysis of RVS; the angle between the major retinal arteries. The variables related to PRAA were identified using model selection with the corrected Akaike information criterion.
Forty-two eyes (26.1%) were allocated to the RVS(+) group. There were seven patterns in the RVS of those in the RVS(+) group, including clockwise shift in the supra temporal area (5 eyes), infra temporal area (7 eyes), and nasal area (9 eyes); anticlockwise shift in the supra temporal area (7 eyes), infra temporal area (5 eyes), and nasal area (2 eyes); and distal shift in the temporal area (7 eyes). The optimal model for the PRAA narrowing included larger AL and body weight in the first grade, and greater AL elongation.
Various (seven) RVS patterns were observed in about 25% of the junior high school students within two years. RVS was associated with AL elongation, and useful to reveal the mechanism of myopic retinal stretch.
Full text
Available for:
DOBA, IZUM, KILJ, NUK, PILJ, PNG, SAZU, SIK, UILJ, UKNU, UL, UM, UPUK