Faculty Bibliography

Iris retraction syndrome (IRS), a potential sign of retinal detachment, is a clinical phenomenon noted in both non-surgical and post-surgical settings. The mechanism of IRS remains unclear and has been attributed to decreased aqueous humor formation and increased fluid clearance through retinal pigment epithelium (RPE) pump action. If IRS is present, prompt pupillary dilation and a thorough exam of the peripheral retina is indicated

PURPOSE: Machine-learning classifiers are trained computerized systems with the ability to detect the relationship between multiple input parameters and a diagnosis. The present study investigated whether the use of machine-learning classifiers improves optical coherence tomography (OCT) glaucoma detection. METHODS: Forty-seven patients with glaucoma (47 eyes) and 42 healthy subjects (42 eyes) were included in this cross-sectional study. Of the glaucoma patients, 27 had early disease (visual field mean deviation [MD] > or = -6 dB) and 20 had advanced glaucoma (MD < -6 dB). Machine-learning classifiers were trained to discriminate between glaucomatous and healthy eyes using parameters derived from OCT output. The classifiers were trained with all 38 parameters as well as with only 8 parameters that correlated best with the visual field MD. Five classifiers were tested: linear discriminant analysis, support vector machine, recursive partitioning and regression tree, generalized linear model, and generalized additive model. For the last two classifiers, a backward feature selection was used to find the minimal number of parameters that resulted in the best and most simple prediction. The cross-validated receiver operating characteristic (ROC) curve and accuracies were calculated. RESULTS: The largest area under the ROC curve (AROC) for glaucoma detection was achieved with the support vector machine using eight parameters (0.981). The sensitivity at 80% and 95% specificity was 97.9% and 92.5%, respectively. This classifier also performed best when judged by cross-validated accuracy (0.966). The best classification between early glaucoma and advanced glaucoma was obtained with the generalized additive model using only three parameters (AROC = 0.854). CONCLUSIONS: Automated machine classifiers of OCT data might be useful for enhancing the utility of this technology for detecting glaucomatous abnormality.

OBJECTIVE: To compare ultrahigh-resolution optical coherence tomography (UHR OCT) with standard-resolution OCT for imaging macular diseases, develop baselines for interpreting OCT images, and identify situations where UHR OCT can provide additional information on disease morphology. DESIGN: Cross-sectional study. PARTICIPANTS: One thousand two eyes of 555 patients with different macular diseases including macular hole, macular edema, central serous chorioretinopathy, age-related macular degeneration (AMD), choroidal neovascularization, epiretinal membrane, retinal pigment epithelium (RPE) detachment, and retinitis pigmentosa. METHODS: A UHR ophthalmic OCT system that achieves 3-microm axial image resolution was developed for imaging in the ophthalmology clinic. Comparative studies were performed with both UHR OCT and standard 10-microm-resolution OCT. Standard scanning protocols of 6 radial 6-mm scans through the fovea were obtained with both systems. Ultrahigh-resolution OCT and standard-resolution OCT images were correlated with standard ophthalmic examination techniques (dilated ophthalmoscopy, fluorescein angiography, indocyanine green angiograms) to assess morphological information contained in the images. MAIN OUTCOME MEASURES: Ultrahigh-resolution and standard-resolution OCT images of macular pathologies. RESULTS: Correlations of UHR OCT images, standard-resolution images, fundus examination, and/or fluorescein angiography were demonstrated in full-thickness macular hole, central serous chorioretinopathy, macular edema, AMD, RPE detachment, epiretinal membrane, vitreal macular traction, and retinitis pigmentosa. Ultrahigh-resolution OCT and standard-resolution OCT exhibited comparable performance in differentiating thicker retinal layers, such as the retinal nerve fiber, inner and outer plexiform, and inner and outer nuclear. Ultrahigh-resolution OCT had improved performance differentiating finer structures or structures with lower contrast, such as the ganglion cell layer and external limiting membrane. Ultrahigh-resolution OCT confirmed the interpretation of features, such as the boundary between the photoreceptor inner and outer segments, which is also visible in standard-resolution OCT. The improved resolution of UHR OCT is especially advantageous in assessing photoreceptor morphology. CONCLUSIONS: Ultrahigh-resolution OCT enhances the visualization of intraretinal architectural morphology relative to standard-resolution OCT. Ultrahigh-resolution OCT images can provide a baseline for defining the interpretation of standard-resolution images, thus enhancing the clinical utility of standard OCT imaging. In addition, UHR OCT can provide additional information on macular disease morphology that promises to improve understanding of disease progression and management.