Ruth Zelkha

Thickness profiles of retinal layers by optical coherence tomography image segmentation.

PURPOSE To report an image segmentation algorithm that was developed to provide quantitative thickness measurement of six retinal layers in optical coherence tomography (OCT) images. DESIGN Prospective cross-sectional study. METHODS Imaging was performed with time- and spectral-domain OCT instruments in 15 and 10 normal healthy subjects, respectively. A dedicated software algorithm was developed for boundary detection based on a 2-dimensional edge detection scheme, enhancing edges along the retinal depth while suppressing speckle noise. Automated boundary detection and quantitative thickness measurements derived by the algorithm were compared with measurements obtained from boundaries manually marked by three observers. Thickness profiles for six retinal layers were generated in normal subjects. RESULTS The algorithm identified seven boundaries and measured thickness of six retinal layers: nerve fiber layer, inner plexiform layer and ganglion cell layer, inner nuclear layer, outer plexiform layer, outer nuclear layer and photoreceptor inner segments (ONL+PIS), and photoreceptor outer segments (POS). The root mean squared error between the manual and automatic boundary detection ranged between 4 and 9 mum. The mean absolute values of differences between automated and manual thickness measurements were between 3 and 4 mum, and comparable to interobserver differences. Inner retinal thickness profiles demonstrated minimum thickness at the fovea, corresponding to normal anatomy. The OPL and ONL+PIS thickness profiles respectively displayed a minimum and maximum thickness at the fovea. The POS thickness profile was relatively constant along the scan through the fovea. CONCLUSIONS The application of this image segmentation technique is promising for investigating thickness changes of retinal layers attributable to disease progression and therapeutic intervention.

Differential Macular Morphology in Patients with RPE65-, CEP290-, GUCY2D-, and AIPL1-Related Leber Congenital Amaurosis

PURPOSE To evaluate genotypic and macular morphologic correlations in patients with RPE65-, CEP290-, GUCY2D-, or AIPL1-related Leber congenital amaurosis (LCA) using spectral-domain optical coherence tomography (SD-OCT). METHODS SD-OCT macular scans were performed in 21 patients, including 10 with RPE65, 7 with CEP290, 3 with GUCY2D, and 1 with AIPL1 mutations. An image processing software was used to manually draw segmentation lines by three observers. Lamellar structure was evaluated based on the number of retinal layers on segmented images. Total retinal thickness was measured at the central macular and perifoveal areas by using an automated algorithm. RESULTS All three patients with GUCY2D mutations (age range, 20-53 years) retained six retinal layers with visible photoreceptor inner/outer segment juncture (PSJ). However, the preservation of lamellar structures did not parallel better visual acuity. Patients with other mutations had poorly defined PSJ and disorganized retinal lamellar structures, where only one to three retinal layers could be observed. Patients with CEP290 mutations trended to have retention of the outer nuclear layer at the fovea and macular thickening, especially at younger ages. In patients with RPE65 (age range, 20-71 years) and AIPL1 mutations (age, 22 years), macular thickness was markedly decreased. Disorganization of retinal lamellar structures in the RPE65 group trended toward a worsening with increasing age. CONCLUSIONS Variations of macular microstructures were observed among LCA patients with different genotypes. Disorganization of retinal lamellar structure was generally age related. Preservation of retinal microanatomic structures may not be associated with better visual acuity.

Feasibility of a Method for En Face Imaging of Photoreceptor Cell Integrity

PURPOSE To report a method for en face imaging of the photoreceptor inner and outer segment junction by spectral-domain optical coherence tomography (SD OCT) and to describe findings in normal subjects and patients with various retinal diseases. DESIGN Observational case series. METHODS SD OCT images were acquired from 6 normal subjects (mean age, 44 ± 11 years) and from 5 subjects with retinal diseases (mean age, 66 ± 22 years). A customized high-density SD OCT volume scan was acquired on the retina. SD OCT B-scan images were segmented automatically to extract intensity data along the inner and outer segment junction. Data obtained from the raster B-scans were combined to generate an inner and outer segment en face image in a 4.4 × 4.4-mm retinal area centered on the fovea. The foveal-to-parafoveal mean intensity ratio was measured, and repeatability was determined. An infrared scanning laser ophthalmoscope image was acquired and was cropped to provide a field of view similar to the inner and outer segment en face image. RESULTS Inner and outer segment en face images generated in normal subjects provided clear visualization of the retinal vasculature, matching the vascular network observed in the infrared scanning laser ophthalmoscope image. In normal subjects, the foveal-to-parafoveal mean intensity ratio was 0.88 ± 0.06, and repeatability of measurements was, on average, 7%. In macular hole, a dark circular region was observed in the inner and outer segment en face image, indicative of photoreceptor cell loss. In age-related macular degeneration, the en face image displayed nonuniform texture corresponding to topographic variations in the inner and outer segment junction. In central serous retinopathy, areas of lower intensity were visible on the en face image corresponding to regions of prior neurosensory elevation. In cystoid macular edema, reduced intensity was present in the inner and outer segment en face image in areas with increased retinal thickness. In diabetic retinopathy, the inner and outer segment en face image displayed regions of reduced intensity resulting from edema, laser scars, or both. CONCLUSIONS Detection of intensity abnormalities in the inner and outer segment en face image is useful for monitoring the integrity of photoreceptor cells in the course of disease progression and therapeutic intervention.

Optical section retinal imaging and wavefront sensing in diabetes.

Purpose. To investigate differences in higher-order ocular aberrations and in optical section retinal image resolution between healthy normal and diabetic subjects. Methods. An optical imaging system was established for combined retinal optical section imaging and wavefront sensing. A laser beam was expanded and focused to a point on the retina by the optics of the eye. For optical section retinal imaging, a cylindrical lens was placed in the path of the incident laser beam to form a focused line on the retina. Because of the angle between the incident laser and imaging path, an optical section image of the retina was captured. For wavefront sensing, a Shack-Hartmann aberrometer was incorporated in the imaging system. Twenty-two subjects with diabetes (average age, 52 ± 12 years) and 13 normal subjects (average age, 47 ± 9 years) were imaged. Retinal depth resolution was determined from the width of the laser line on the retina. Higher-order ocular aberrations were determined from the root mean square of the third to seventh Zernike terms, characterizing the wavefront aberration function. The data were analyzed statistically using Student’s t-test and linear regression. Results. Higher-order ocular aberrations in diabetic subjects were significantly higher than in normal subjects (p = 0.03). The retinal image depth resolution in diabetic subjects was significantly lower than in normal subjects (p < 0.001). The retinal image depth resolution was inversely correlated with higher-order aberrations (r = −0.5; p = 0.007; N = 35). Conclusions. The results demonstrate disease-related increases in higher-order ocular aberrations that influence retinal image resolution in diabetic eyes. This information is useful for designing high-resolution retinal imaging systems applicable for eyes with retinal disease.

Induction of autophagy in rats upon overexpression of wild-type and mutant optineurin gene

BackgroundOptineurin is a gene associated with normal tension glaucoma and amyotrophic lateral sclerosis. It has been reported previously that in cultured RGC5 cells, the turnover of endogenous optineurin involves mainly the ubiquitin-proteasome pathway (UPP). When optineurin is upregulated or mutated, the UPP function is compromised as evidenced by a decreased proteasome β5 subunit (PSMB5) level and autophagy is induced for clearance of the optineurin protein.ResultsAdeno-associated type 2 viral (AAV2) vectors for green fluorescence protein (GFP) only, GFP-tagged wild-type and Glu50Lys (E50K) mutated optineurin were intravitreally injected into rats for expression in retinal ganglion cells (RGCs). Following intravitreal injections, eyes that received optineurin vectors exhibited retinal thinning, as well as RGC and axonal loss compared to GFP controls. By immunostaining and Western blotting, the level of PSMB5 and autophagic substrate degradation marker p62 was reduced, and the level of autophagic marker microtubule associated protein 1 light chain 3 (LC3) was enhanced. The UPP impairment and autophagy induction evidently occurred in vivo as in vitro. The optineurin level, RGC and axonal counts, and apoptosis in AAV2-E50K-GFP-injected rat eyes were averted to closer to normal limits after treatment with rapamycin, an autophagic enhancer.ConclusionsThe UPP function was reduced and autophagy was induced when wild-type and E50K optineurin was overexpressed in rat eyes. This study validates the in vitro findings, confirming that UPP impairment and autophagy induction also occur in vivo. In addition, rapamycin is demonstrated to clear the accumulated mutant optineurin. This agent may potentially be useful for rescuing of the adverse optineurin phenotypes in vivo.

Conjunctival microvascular haemodynamics in sickle cell retinopathy.

To determine alterations in bulbar conjunctival microvascular haemodynamics in sickle cell retinopathy (SCR) subjects with focal macular thinning (FMT).

Nerve fiber layer thickness in glaucoma patients with asymmetric hemifield visual field loss.

PurposeTo investigate the presence of retinal nerve fiber layer (RNFL) thinning and determine the relationship between RNFL thickness and visual field sensitivity loss in glaucoma patients with asymmetric hemifield visual field loss. Patients and MethodsThirty glaucoma patients with asymmetric hemifield visual field loss and 30 normal control subjects were included in the study. RNFL thickness was measured by optical coherence tomography and visual field sensitivity was measured by automated perimetry. Glaucoma patients with advanced visual field loss restricted to 1 hemifield and early or absent glaucomatous field loss in the other hemifield on the basis of the visual field data were included. Visual field sensitivity and mean deviation (MD) were averaged separately in each of the 2 hemifields. The hemifields in each eye were categorized as early (MD≥−6 dB) and advanced (MD<−6 dB) glaucomatous hemifields. ResultsRNFL thickness measurements in corresponding (eg, superior peripapillary quadrant with inferior hemifield) advanced glaucomatous hemifields (59±16 μm) were significantly (P<0.001) lower than in corresponding early glaucomatous hemifields (90±25 μm). The mean RNFL thickness in corresponding advanced and early glaucomatous hemifields were significantly lower than in normal control subjects (P<0.0001). On the basis of the normative database supplied by optical coherence tomography software, 100% and 43% of eyes had abnormal RNFL thickness in corresponding advanced and early glaucomatous hemifields, respectively. A linear correlation was found between RNFL thickness and MD in the early (r=0.6; P<0.001) and advanced (r=0.5; P=0.007) glaucomatous hemifields. ConclusionsRNFL thinning was present in corresponding hemifields of glaucomatous eyes with minimal visual field defect and correlated with visual field sensitivity loss. Measurement of RNFL thickness has potential for detection of early nerve fiber loss owing to glaucoma.

The relationship between macular cell layer thickness and visual function in different stages of glaucoma

PurposeTo determine whether there were differences in the structure–function relationship between early and advanced glaucoma, and study the association between thickness of discrete macular cell layers, the thickness of the retinal nerve fiber layer, and visual field sensitivity.MethodsIn all, 71 eyes of 50 subjects (28 glaucoma patients and 22 normal control subjects) were included. Thickness of macular retinal nerve fiber layer (mRNFL), macular inner retinal layer (mIRL), and macular outer retinal layer (mORL) were measured from Stratus optical coherence tomography macular scans, using our previously published segmentation algorithm. Visual sensitivity loss was determined by mean deviation (MD) using Humphrey Visual Field Analyzer. The mean thickness for each layer from the normal control subjects, early, and advanced glaucoma groups was compared. In addition, a mixed model analysis was used to explore the relationship between structure–function, allowing for possible interaction with glaucoma stage.ResultsThe mean mRNFL thickness in early and advanced glaucoma patients was significantly less than measurements in normal subjects (P<0.01). The mean mIRL thickness in advanced glaucoma was significantly less than normal subjects (P=0.04). The mean mORL thickness in early and advanced glaucoma was not statistically significant different from that of normal subjects (P>0.8). There was no statistically significant difference in macular structure–function relationship between the two glaucoma groups (P>0.05). Mean mIRL thickness was significantly associated with MD (P=0.04).ConclusionThere was no significant difference in macular structure–function relationship between early and advanced glaucoma groups. Combined data from both glaucoma groups indicated that mIRL thickness was associated with visual sensitivity loss.

EN FACE SPECTRAL-DOMAIN OPTICAL COHERENCE TOMOGRAPHY OUTER RETINAL ANALYSIS AND RELATION TO VISUAL ACUITY

Purpose: To describe a method of en face visualization and quantification of the photoreceptor inner segment/outer segment junction area, using spectral-domain optical coherence tomography, and association with visual acuity. Methods: Case series of 74 eyes in 53 patients. Central 1-mm and 400-&mgr;m en face areas were analyzed with a computer algorithm. Results: The presence or absence of inner segment/outer segment junction was visible on both spectral-domain optical coherence tomography en face and retinal cross sections. Thirty eyes (40.6%) had no retinal pathology and an average logMAR visual acuity of 0.116. Twenty-five eyes (33.8%) had intraretinal edema, with visual acuity of 0.494. Nineteen eyes had nonneovascular age-related macular degeneration (dry age-related macular degeneration, 25.6%), with visual acuity of 0.392. In all eyes, central 1-mm and 400-&mgr;m en face areas were 58.3 ± 25.0% and 56.4 ± 26.0%, which showed significant correlation with visual acuity (Pearson correlation, r = −0.66 and −0.56, both P < 0.001). This correlation was greater than correlation of visual acuity with central subfield thickness (r = 0.39, P < 0.001), macular volume (r = 0.36, P = 0.002), and average macular thickness (r = 0.37, P = 0.001). However, no variables were significantly correlated with dry age-related macular degeneration eyes. Conclusion: Central en face inner segment/outer segment junction areas are significantly correlated with visual acuity in most eyes. This may correlate better with visual acuity than other spectral-domain optical coherence tomography values, as a reflection of photoreceptor integrity. Dry age-related macular degeneration may disrupt the plane used to formulate the en face display. Advancements in spectral-domain optical coherence tomography may provide routine en face visualization analysis.

Feasibility of Noninvasive Imaging of Chorioretinal Oxygenation

BACKGROUND AND OBJECTIVE The feasibility of an optical system for noninvasive imaging of chorioretinal oxygenation was evaluated. Due to its depth discrimination, this optical section imaging technique has potential for differential imaging of oxygen tension in the chorioretinal vasculatures. MATERIALS AND METHODS The method consisted of projecting a narrow laser line obliquely on the retina after intravenous injection of an oxygen-sensitive probe and imaging the phosphorescence emission. Due to the angle between the incident laser and imaging path, a phosphorescence optical section image of the retina was captured. The phosphorescence intensity was measured in the chorioretinal vasculatures. The method was tested in three rats while breathing 10% oxygen, 50% oxygen, and room air. RESULTS On the phosphorescence optical section image, vasculatures appeared laterally displaced according to their depth location, displaying probe phosphorescence separately in the chorioretinal vasculatures. Oxygenation increased in all vasculatures with increased inhaled percent oxygen. Oxygenation in the retinal artery was 2.3, 1.9, and 1.6 times oxygenation in the retinal vein, capillary, and choroid, respectively. During hypoxia, oxygenation decreased by 28%, 18%, 22%, and 14% in the retinal vein, artery, capillary, and choroid, respectively. During hyperoxia, oxygenation increased by 30%, 45%, 36%, and 28% in the retinal vein, artery, capillary, and choroid, respectively. CONCLUSION The results demonstrate the feasibility of this technique for noninvasive and separate imaging of chorioretinal oxygenation and its potential for three-dimensional oxygen tension imaging.