Optical Coherence Tomography Thickness Maps Research Articles

SLO-Net: Enhancing Multiple Sclerosis Diagnosis Beyond Optical Coherence Tomography Using Infrared Reflectance Scanning Laser Ophthalmoscopy Images.

Several machine learning studies have used optical coherence tomography (OCT) for multiple sclerosis (MS) classification with promising outcomes. Infrared reflectance scanning laser ophthalmoscopy (IR-SLO) captures high-resolution fundus images, commonly combined with OCT for fixed B-scan positions. However, no machine learning research has utilized IR-SLO images for automated MS diagnosis. This study utilized a dataset comprised of IR-SLO images and OCT data from Isfahan, Iran, encompassing 32 MS and 70 healthy individuals. A number of convolutional neural networks (CNNs)-namely, VGG-16, VGG-19, ResNet-50, ResNet-101, and a custom architecture-were trained with both IR-SLO images and OCT thickness maps as two separate input datasets. The highest performing models for each modality were then integrated to create a bimodal model that receives the combination of OCT thickness maps and IR-SLO images. Subject-wise data splitting was employed to prevent data leakage among training, validation, and testing sets. Overall, images of the 102 patients from the internal dataset were divided into test, validation, and training subsets. Subsequently, we employed a bootstrapping approach on the training data through iterative sampling with replacement. The performance of the proposed bimodal model was evaluated on the internal test dataset, demonstrating an accuracy of 92.40% ± 4.1% (95% confidence interval [CI], 83.61-98.08), sensitivity of 95.43% ± 5.75% (95% CI, 83.71-100.0), specificity of 92.82% ± 3.72% (95% CI, 81.15-96.77), area under the receiver operating characteristic (AUROC) curve of 96.99% ± 2.99% (95% CI, 86.11-99.78), and area under the precision-recall curve (AUPRC) of 97.27% ± 2.94% (95% CI, 86.83-99.83). Furthermore, to assess the model generalization ability, we examined its performance on an external test dataset following the same bootstrap methodology, achieving promising results, with accuracy of 85.43% ± 0.08% (95% CI, 71.43-100.0), sensitivity of 97.33% ± 0.06% (95% CI, 83.33-100.0), specificity of 84.6% ± 0.10% (95% CI, 71.43-100.0), AUROC curve of 99.67% ± 0.02% (95% CI, 95.63-100.0), and AUPRC of 99.65% ± 0.02% (95% CI, 94.90-100.0). Incorporating both modalities improves the performance of automated diagnosis of MS, showcasing the potential of utilizing IR-SLO as a complementary tool alongside OCT. Should the results of our proposed bimodal model be validated in future work with larger and more diverse datasets, diagnosis of MS based on both OCT and IR-SLO can be reliably integrated into routine clinical practice.

Segmentation-Free OCT-Volume-Based Deep Learning Model Improves Pointwise Visual Field Sensitivity Estimation.

The structural changes measured by optical coherence tomography (OCT) are related to functional changes in visual fields (VFs). This study aims to accurately assess the structure-function relationship and overcome the challenges brought by the minimal measurable level (floor effect) of segmentation-dependent OCT measurements commonly used in prior studies. We developed a deep learning model to estimate the functional performance directly from three-dimensional (3D) OCT volumes and compared it to the model trained with segmentation-dependent two-dimensional (2D) OCT thickness maps. Moreover, we proposed a gradient loss to utilize the spatial information of VFs. Our 3D model was significantly better than the 2D model both globally and pointwise regarding both mean absolute error (MAE = 3.11 + 3.54 vs. 3.47 ± 3.75 dB, P < 0.001) and Pearson's correlation coefficient (0.80 vs. 0.75, P < 0.001). On a subset of test data with floor effects, the 3D model showed less influence from floor effects than the 2D model (MAE = 5.24 ± 3.99 vs. 6.34 ± 4.58 dB, P < 0.001, and correlation 0.83 vs. 0.74, P < 0.001). The gradient loss improved the estimation error for low-sensitivity values. Furthermore, our 3D model outperformed all prior studies. By providing a better quantitative model to encapsulate the structure-function relationship more accurately, our method may help deriving VF test surrogates. DL-based VF surrogates not only benefit patients by reducing the testing time of VFs but also allow clinicians to make clinical judgments without the inherent limitations of VFs.

Retinal Vessel Density in Age-Related Macular Degeneration Patients with Geographic Atrophy

We compared the retinal vessel density and inner retinal thickness in patients who had one eye with geographic atrophy (GA) and a fellow eye with intermediate age-related macular degeneration (iAMD). The vessel density from the superficial vascular complex (SVC) and deep vascular complex (DVC) through optical coherence tomography angiography and the thickness of the nerve fiber layer, ganglion cell–inner plexiform layer (GCIPL), inner nuclear layer (INL), outer nuclear layer (ONL) on a structural optical coherence tomography thickness map were measured in 28 eyes of 14 GA patients with iAMD in the fellow eye. GA eyes had significantly lower vessel density in the SVC (26.2 ± 3.9% vs. 28.3 ± 4.4%; p = 0.015) and DVC (24.2 ± 2.6% vs. 26.8 ± 1.9%; p = 0.003) than fellow eyes (iAMD). GCIPL and ONL were significantly thinner in GA eyes than in the fellow eyes (p = 0.032 and 0.024 in the foveal areas, p = 0.029 and 0.065 in the parafovea areas, respectively). Twenty-four eyes of 12 patients were followed up for 2 years and seven of the fellow eyes (58.3%) developed GA during the follow-up period and showed reduced vessel density in the SVC (26.4 ± 3.0% vs. 23.8 ± 2.9%; p = 0.087) and DVC (25.8 ± 2.2% vs. 22.4 ± 4.4%; p = 0.047) compared to baseline. Vessel density and GCIPL thickness map measurements are potential GA markers in non-neovascular AMD.

Effect of Anti-glaucoma Agents on Myopic Retinoschisis.

PurposeTo evaluate the effect of intraocular pressure (IOP)-lowering medications on myopic retinoschisis.MethodsThe medical records of 33 patients (36 eyes) with myopic retinoschisis associated with pathologic myopia were reviewed retrospectively. The patients were divided into two groups: the study group comprising patients undergoing treatment with anti-glaucoma medications for suspected glaucoma; the control group comprising patients who did not use any IOP lowering medications. The changes in retinoschisis in the two groups were compared using the Spectralis domain optical coherence tomography thickness map protocol.ResultsThe study group included 18 eyes (17 patients), and the control group included 18 eyes (16 patients). There were no significant differences between the 6-month and 12-month improvement or aggravation rates of the two groups (p = 0.513 and 0.137, respectively). However, after 18 months, the aggravation rate of retinoschisis was significantly lower in the study group (p = 0.003). The improvement / aggravation rate was 58.33% / 16.67% in the study group and 0% / 57.14% in the control group.ConclusionsThe use of IOP-lowering medications for more than a year may be useful for the management of retinoschisis associated with pathologic myopia.

Paediatric optical coherence tomography normative databases: A real need

Abstract Introduction Optical coherence tomography (OCT) has become an essential tool in paediatric ophthalmology. However, none of the currently available OCT devices include any kind of normative database for the paediatric population, which can lead to important interpretation errors. Objective To review the paediatric OCT normative databases. Material and methods The applications and implications of the use of paediatric OCT normative databases are reviewed. Results The paediatric normative databases that have been published so far in scientific literature with different OCT devices for Spanish and European population are presented. Conclusions The knowledge and interpretation of paediatric OCT normative databases in our daily clinical practice is crucial in order for the correct interpretation of OCT thickness maps.

Serial Combined Wide-Field Optical Coherence Tomography Maps for Detection of Early Glaucomatous Structural Progression

Both parapapillary and macular areas are important in determining the progression of early glaucoma. However, no attempt has been made to assess the progression of glaucoma in images that combine the 2 areas. To evaluate the potential usefulness of serial analysis of combined wide-field optical coherence tomography (OCT) maps for detection of structural progression in patients with early glaucoma. Retrospective observational study. Patients with early primary open-angle glaucoma with a minimum of 3-year follow-up involving serial spectral-domain OCT measurement were analyzed. Patients were divided into a nonprogressor group (n = 47) and a progressor group (n = 47) on the basis of serial stereo disc photography and red-free photography. Serial combined wide-field OCT maps integrating parapapillary retinal nerve fiber layer (RNFL) and macular ganglion cell-inner plexiform layer (GCIPL) maps were generated with the embedded software of serial spectral-domain OCT. Glaucoma specialists then assessed the structural progression detection ability of those serial wide-field OCT maps for early glaucomatous eyes and compared their sensitivity with those of RNFL and GCIPL guided progression analyses (GPAs). The diagnostic ability of the serial wide-field OCT maps for early glaucomatous structural progression. Ninety-four patients (mean [SD] age, 51.4 [12.3] years; 48 [51.1%] women; all Korean) were included. The serial wide-field OCT map analysis showed good agreement for detection of structural progression between the 2 glaucoma graders (wide-field OCT thickness map: κ = 0.649; wide-field OCT deviation map: κ = 0.833). These maps showed early glaucomatous structural progression detection abilities comparable with those of RNFL and GCIPL GPAs (sensitivities of wide-field OCT thickness map, wide-field OCT deviation map, RNFL GPA, and GCIPL GPA = 63.8%, 83.0%, 83.0%, and 66.0%, respectively, all P > .05; specificities of wide-field OCT thickness map, wide-field OCT deviation map, RNFL GPA, and GCIPL GPA = 93.6%, 95.7%, 84.8%, and 93.6%, respectively, all P > .05). The serial combined wide-field OCT maps integrating RNFL and GCIPL maps performed well in detecting structural progression in early glaucomatous eyes. Confirmation in an independent prospective study might provide greater confidence in this conclusion.

Computer‐aided analysis of fluorescein angiograms using colour leakage maps

Fundus fluorescein angiography (FFA) is a standard screening and diagnosis technique for several retinal diseases. The analysis of FFA images is performed qualitatively by skilled observers, and thus is vulnerable to inter- and intra-observer variability. In this study, the authors present a method for computer-aided analysis of FFA images. The method is based on generating quantitative colour fluorescein leakage maps (FLM) that mimic the thickness maps generated by the optical coherence tomography (OCT). Results from 64 patients show strong correlation between the FLM and OCT thickness maps (r = 0.8). The method was found to be reproducible and robust to variability in the image acquisition times.

Effect of Grid Decentration on Macular Thickness Measurements in Normal Subjects and Patients with Diabetic Macular Edema

To evaluate the effect of optical coherence tomography (OCT) grid decentration on macular thickness measurements in healthy subjects and patients with diabetic macular edema. In this prospective study, 3D spectral-domain OCT images of 51 eyes with clinically significant diabetic macular edema and 29 healthy eyes were assessed. In each eye, the macular Early Treatment Diabetic Retinopathy Study grid was evaluated for decentration. After grid adjustment, changes in central subfield thickness (CST) and central subfield volume (CSV) measurements were recorded. Changes &gt;1 μm and &gt;8.5 μm in CST were considered as grid decentration and clinically significant grid decentration, respectively. Grid decentration was found in 10 normal eyes (34.6%) and 32 clinically significant macular edema (CSME) eyes (62.7%, p = 0.01). Clinically significant CST changes were found in 3 normal eyes (10.3%) and 23 CSME eyes (45%, p = 0.001). The mean change in CST after grid adjustment was 6.4 ± 5.8 µm and 24.7 ± 27 µm in normal and CSME eyes (p = 0.04). Age, sex, and CST had no statistically significant effect on grid decentration in normal and CSME eyes. Best-corrected visual acuity was significantly worse in eyes with grid decentration (0.75 ± 0.55 vs 0.42 ± 0.22 logMAR, p = 0.02) and clinically significant grid decentration (0.81 ± 0.6 vs 0.45 ± 0.27 logMAR, p = 0.01). Change in CSV was detected in 3 normal eyes (10.3%) and 24 CSME eyes (47%, p = 0.001). Grid decentration occurs in a large number of OCT thickness maps and leads to significant error in macular thickness measurements. The effect is more prominent in patients with CSME.

Agreement of Retinal Nerve Fiber Layer Defect Location Between Red-Free Fundus Photography and Cirrus HD-OCT Maps

Purpose: To investigate the agreement of angular locations of retinal nerve fiber layer (RNFL) defect margins in glaucomatous eyes by using red-free fundus photographs and Cirrus high-definition optical coherence tomography (OCT) RNFL deviation and thickness maps.Methods: We examined 380 RNFL defects that showed clear margins in red-free fundus photographs. The OCT deviation and thickness maps were overlaid on the corresponding red-free fundus photographs. A reference line was drawn between the disc center and the macular center. Lines were also drawn between the optic disc center and the point where the RNFL defect margins crossed the OCT scan circle. The angle between the reference and defect-margin lines defined the angular location of the defect margin. Angular locations of proximal (nearest to the reference) and distal (farthest from the reference) RNFL defect margins on OCT deviation and thickness maps were compared to the locations on red-free fundus photographs.Results: The angular locations of proximal and distal RNFL defect margins on OCT thickness maps showed good agreement with red-free fundus photographs. However, OCT deviation maps showed greater angular locations for both proximal and distal RNFL defect margins compared with red-free fundus photographs, especially in eyes with higher myopia (p < 0.05).Conclusions: Red-free fundus photographs and OCT thickness maps showed good agreement for the RNFL defect margin identification. However, this was not the case for deviation maps, especially in myopic eyes. This finding should be considered when evaluating RNFL defects using OCT maps.

Retinal Nerve Fiber Layer Progression in Glaucoma: A Comparison between Retinal Nerve Fiber Layer Thickness and Retardance

To investigate the performance of spectral-domain optical coherence tomography (OCT) and scanning laser polarimetry to detect progressive retinal nerve fiber layer (RNFL) changes and to determine whether reduction of the RNFL retardance occurred before thinning of the RNFL in glaucoma. Prospective, longitudinal study. One hundred eighty-four eyes of 116 glaucoma patients and 43 normal eyes of 23 healthy individuals. Patients were followed up every 4 months for at least 36 months with RNFL retardance (GDx Enhanced Corneal Compensation; Carl Zeiss Meditec) and RNFL thickness (Cirrus HD-OCT; Carl Zeiss Meditec) measured in the same visit. Progressive RNFL retardance and thickness changes were evaluated with event-based analysis (Guided Progression Analysis; Carl Zeiss Meditec) with reference to the RNFL retardance change map and the RNFL thickness change map, respectively. The area and frequency distribution of RNFL changes were examined by overlaying the RNFL retardance change maps and the RNFL thickness change maps in the latest follow-up. The agreement of RNFL retardance and RNFL thickness progression was evaluated with κ statistics. Number of eyes with progressive RNFL changes over time. A total of 2472 OCT thickness maps and 2472 RNFL retardance maps were collected and reviewed with a mean follow-up of 55.1 months. Twenty-seven eyes (14.6%; 26 glaucoma patients) showed progressive RNFL thinning, whereas 8 eyes (4.3%; 8 glaucoma patients) showed progressive reduction of RNFL retardance. Seven eyes (3.8%; 7 glaucoma patients) had progression that was detected by both instruments, all with progressive RNFL thinning detected before progressive reduction of RNFL retardance became evident, and the mean lag time was 13.4 months (range, 4.0-37.6 months). The agreement between RNFL thickness and RNFL retardance progression was fair (κ, 0.357). Progressive loss of RNFL thickness was observed most frequently at the inferotemporal 223° to 260°, whereas the inferotemporal 227° to 263° and superior 56° to 117° were observed most commonly for progressive loss of RNFL retardance. In the normal group, no eyes showed reduction in RNFL thickness or retardance. At a comparable level of specificity, progressive RNFL thinning was detected more often than progressive reduction of RNFL retardance. For eyes with progressive loss of RNFL thickness and RNFL retardance, the former preceded the latter.

Functional and morphological changes in diabetic macular edema over the course of anti-vascular endothelial growth factor treatment

To evaluate macular morphology and function in diabetic macular edema (DME) over the course of intravitreal anti-vascular endothelial growth factor (VEGF) treatment with Ranibizumab. A consecutive series of 39 study eyes with centre-involving DME were included in this study. In all subjects, best-corrected visual acuity (BCVA) according ETDRS protocol, fluorescein angiography (FA), microperimetric macular sensitivity (MP) and Spectral Domain optical coherence tomography (SD-OCT) cross-sectional scans were obtained before treatment and after 3 monthly applied intravitreal Ranibizumab injections. Six different morphological qualities [IS/OS layer integrity, outer nuclear layer (ONL) cysts, ONL cyst size, inner nuclear layer (INL) cysts, blocking phenomenon and subretinal fluid] were graded of each cross-sectional OCT scan before and over the course of treatment by two experienced graders. Correlation analyses between functional and morphological parameters were obtained. Mean BCVA increased from 26 ± 14 to 33 ± 13 letters after 3 consecutive monthly applied Ranibizumab injections (p < 0.001). Central retinal thickness (CRT) decreased from 504 ± 144 to 387 ± 122 μm (p < 0.001). Over the course of treatment, IS/OS continuity improved (index: 0.56 ± 0.52 to 0.43 ± 0.49, Z = -1.415, p = 0.157), ONL cyst prevalence and size decreased significantly (index: 0.61 ± 0.44 to 0.56 ± 0.35, Z = -3.41, p = 0.001 and 1.75 ± 0.88 to 1.17 ± 1.05, Z = -4.02, p < 0.001), INL cyst prevalence decreased (index: 0.35 ± 0.52 to 0.28 ± 0.52, Z = -1.60, p = 0.109), blocking phenomenon did not change significantly (index: 00.12 ± 0.16 to 0.13 ± 0.15, Z = -0.45, p = 0.656) and subretinal fluid almost disappeared (index: 0.10 ± 0.24 vs. 0.00 ± 0.01, Z = -2.56, p = 0.011). Correlation analyses revealed highest significant correlations between ONL cyst prevalence and their size and CRT as well as BCVA and MP before treatment and over the course of treatment. ONL cysts and their size as morphological parameters correlate with retinal function measured with BCVA and microperimetry before and over the course of anti-VEGF therapy with Ranibizumab in patients with DME.