Priyatham S. Mettu

Fully automated detection of diabetic macular edema and dry age-related macular degeneration from optical coherence tomography images

We present a novel fully automated algorithm for the detection of retinal diseases via optical coherence tomography (OCT) imaging. Our algorithm utilizes multiscale histograms of oriented gradient descriptors as feature vectors of a support vector machine based classifier. The spectral domain OCT data sets used for cross-validation consisted of volumetric scans acquired from 45 subjects: 15 normal subjects, 15 patients with dry age-related macular degeneration (AMD), and 15 patients with diabetic macular edema (DME). Our classifier correctly identified 100% of cases with AMD, 100% cases with DME, and 86.67% cases of normal subjects. This algorithm is a potentially impactful tool for the remote diagnosis of ophthalmic diseases.

Kernel regression based segmentation of optical coherence tomography images with diabetic macular edema

We present a fully automatic algorithm to identify fluid-filled regions and seven retinal layers on spectral domain optical coherence tomography images of eyes with diabetic macular edema (DME). To achieve this, we developed a kernel regression (KR)-based classification method to estimate fluid and retinal layer positions. We then used these classification estimates as a guide to more accurately segment the retinal layer boundaries using our previously described graph theory and dynamic programming (GTDP) framework. We validated our algorithm on 110 B-scans from ten patients with severe DME pathology, showing an overall mean Dice coefficient of 0.78 when comparing our KR + GTDP algorithm to an expert grader. This is comparable to the inter-observer Dice coefficient of 0.79. The entire data set is available online, including our automatic and manual segmentation results. To the best of our knowledge, this is the first validated, fully-automated, seven-layer and fluid segmentation method which has been applied to real-world images containing severe DME.

Retinal Artery-Vein Classification via Topology Estimation

We propose a novel, graph-theoretic framework for distinguishing arteries from veins in a fundus image. We make use of the underlying vessel topology to better classify small and midsized vessels. We extend our previously proposed tree topology estimation framework by incorporating expert, domain-specific features to construct a simple, yet powerful global likelihood model. We efficiently maximize this model by iteratively exploring the space of possible solutions consistent with the projected vessels. We tested our method on four retinal datasets and achieved classification accuracies of 91.0%, 93.5%, 91.7%, and 90.9%, outperforming existing methods. Our results show the effectiveness of our approach, which is capable of analyzing the entire vasculature, including peripheral vessels, in wide field-of-view fundus photographs. This topology-based method is a potentially important tool for diagnosing diseases with retinal vascular manifestation.

Retinal thickening in iridocyclitis.

PURPOSE To determine the frequency of retinal thickening (RT) in eyes with iridocyclitis and to examine the correlations among anterior chamber (AC) inflammation, RT, and visual acuity. DESIGN Retrospective, observational case series. METHODS Records were reviewed of patients seen at the Duke University Eye Center Uveitis Clinic from January 2002 through March 2008. Patients with unilateral iridocyclitis without panuveitis, vitreitis, pars planitis, posterior uveitis, or a combination thereof who had undergone optical coherence tomography (OCT) of both eyes were included. The AC cell grade and OCT-RT measurements were recorded. Subretinal fluid and intraretinal cysts were determined from OCT scans according to Duke Reading Center guidelines. RESULTS Forty-three patients were studied. RT typically was present in a ring-like distribution around the fovea. The median difference between the study eye and fellow eye in RT was statistically significant for total macular volume (TMV) and for all OCT subfields (P < .001). In the study eye, there was a modest correlation between the RT and AC cell grade for the OCT-TMV (P = .039; r(2) = 0.1) and the subfield comprised of the quadrants in the outer ring on OCT (P = .027; r(2) = 0.12), and between RT and visual acuity for OCT-TMV and all but the central subfields (P = .003 to .007; r(2) = 0.261 to 0.227). RT decreased after anti-inflammatory therapy. CONCLUSIONS RT is strongly associated with iridocyclitis and decreases after treatment. RT, as determined by OCT, is a useful clinical parameter to evaluate patients with iridocyclitis and to monitor response to treatment.

Image Inversion Spectral-Domain Optical Coherence Tomography Optimizes Choroidal Thickness and Detail through Improved Contrast

PURPOSE This study was conducted to determine whether there were significant differences in choroidal thickness, contrast, outer choroidal vessel (OCV), and choroidal-scleral junction (CSJ) visualization in inverted versus upright spectral-domain optical coherence tomography (SD-OCT). METHODS Images were captured on Bioptigen SD-OCT, Zeiss Cirrus HD-OCT, and Heidelberg Spectralis in 42 eyes of 21 healthy subjects. Average choroidal thickness across a fovea-centered 4-mm segment was determined with MATLAB. Quantitative measures of choroidal contrast were measured and CSJ assessed by applying a score of 0 to 3. OCV was determined by counting choroidal vessels ≥ 200 μm. RESULTS Mean choroidal thickness was greater in inverted versus upright images captured by Bioptigen (P ≤ 0.003) and Spectralis (P ≤ 0.015). Choroidal thickness varied significantly between the three machines (P < 0.05). Contrast was higher in inverted versus upright images captured by Bioptigen (P ≤ 0.02) and Spectralis (P < 0.001), but not in Cirrus (P > 0.10, both observers). CSJ score was highest in the following: Spectralis inverted = Spectralis EDI > Cirrus upright > Bioptigen inverted. Mean OCV was highest in Spectralis inverted mode. CONCLUSION The most favorable modes to visualize CSJ and OCV are the Spectralis EDI, Spectralis inverted, Cirrus upright, and Bioptigen inverted. These modes demonstrate the highest outer choroidal contrast and choroidal thickness measurements. Choroidal thickness cannot be compared between machines due to conversion factor differences. Future studies and construction of automated segmentation and detection software should take these benefits and pitfalls into account.

Fully automatic segmentation of fluorescein leakage in subjects with diabetic macular edema.

PURPOSE To create and validate software to automatically segment leakage area in real-world clinical fluorescein angiography (FA) images of subjects with diabetic macular edema (DME). METHODS Fluorescein angiography images obtained from 24 eyes of 24 subjects with DME were retrospectively analyzed. Both video and still-frame images were obtained using a Heidelberg Spectralis 6-mode HRA/OCT unit. We aligned early and late FA frames in the video by a two-step nonrigid registration method. To remove background artifacts, we subtracted early and late FA frames. Finally, after postprocessing steps, including detection and inpainting of the vessels, a robust active contour method was utilized to obtain leakage area in a 1500-μm-radius circular region centered at the fovea. Images were captured at different fields of view (FOVs) and were often contaminated with outliers, as is the case in real-world clinical imaging. Our algorithm was applied to these images with no manual input. Separately, all images were manually segmented by two retina specialists. The sensitivity, specificity, and accuracy of manual interobserver, manual intraobserver, and automatic methods were calculated. RESULTS The mean accuracy was 0.86 ± 0.08 for automatic versus manual, 0.83 ± 0.16 for manual interobserver, and 0.90 ± 0.08 for manual intraobserver segmentation methods. CONCLUSIONS Our fully automated algorithm can reproducibly and accurately quantify the area of leakage of clinical-grade FA video and is congruent with expert manual segmentation. The performance was reliable for different DME subtypes. This approach has the potential to reduce time and labor costs and may yield objective and reproducible quantitative measurements of DME imaging biomarkers.

Bone Marrow Transplantation Transfers Age-Related Susceptibility to Neovascular Remodeling in Murine Laser- Induced Choroidal Neovascularization

PURPOSE Neovascular remodeling (NVR), the progression of small capillaries into large-caliber arterioles with perivascular fibrosis, represents a major therapeutic challenge in neovascular age-related macular degeneration (AMD). Neovascular remodeling occurs after laser-induced choroidal neovascularization (CNV) in aged but not young mice. Additionally, bone marrow-derived cells, including macrophages, endothelial precursor cells, and mesenchymal precursor cells, contribute to CNV severity. In this study, we investigated the impact of aged bone marrow transplantation (BMT) on the degree of fibrosis, size, and vascular morphology of CNV lesions in a mouse model of laser-induced CNV. METHODS Young (2 months) and old (16 months) mice were transplanted with green fluorescent protein (GFP)-labeled bone marrow isolated from either young or old donors. Laser CNV was induced 1 month following transplant, and eyes were analyzed via choroidal flat mounts and immunohistochemistry 1 month postlaser. The identity of cells infiltrating CNV lesions was determined using specific markers for the labeled transplanted cells (GFP+), macrophages (F4/80+), perivascular mesenchymal-derived cells (smooth muscle actin, SMA+), and endothelial cells (CD31+). RESULTS Bone marrow transplantation from aged mice transferred susceptibility to NVR into young recipients. Inversely, transplantation of young marrow into old mice prevented NVR, preserving small size and minimal fibrosis. Mice with NVR demonstrated a greater relative contribution of marrow-derived SMA+ perivascular mesenchymal cells as compared to other cells. CONCLUSIONS Our findings indicate that the status of bone marrow is an important determining factor of neovascular severity. Furthermore, we find that perivascular mesenchymal cells, rather than endothelial cells, derived from aged bone marrow may contribute to increased CNV severity in this murine model of experimental neovascularization.

Semiautomatic Segmentation of Rim Area Focal Hyperautofluorescence Predicts Progression of Geographic Atrophy Due to Dry Age-Related Macular Degeneration.

Purpose To develop image analysis software usable by nonexpert graders to segment geographic atrophy (GA) from dry AMD and to quantify rim area focal hyperautofluorescence (RAFH) surrounding GA on fundus autofluorescence (FAF) images. To compare the GA progression predictions based on RAFH with those of a validated qualitative classification system. Methods Retrospective analysis of serial FAF images from 49 eyes of 30 subjects with GA was performed using MATLAB-based software (MathWorks, Natick, MA, USA). Correlation between RAFH and progression of GA was analyzed using Spearman correlation. Comparisons of lesion growth rate between RAFH tertiles used generalized estimating equations and Kruskal-Wallis testing. Interobserver variability in lesion size, growth rate and RAFH were compared between two expert and one nonexpert grader using Bland-Altman statistics. Results Rim area focal hyperautofluorescence was positively correlated with GA progression rate (ρ = 0.49, P < 0.001). Subjects in the middle or highest RAFH tertile were at greater risk of progression (P = 0.005 and P = 0.001, respectively). Mean difference in RAFH was 0.012 between expert and −0.005 to 0.017 between expert and nonexperts. Mean difference in lesion size (mm2) was 0.11 between expert and −0.29 to 0.41 between expert and nonexperts. Mean difference in lesion growth rate (mm2/mo) was 0.0098 between expert and −0.027 to 0.037 between expert and nonexperts. Risk stratification based on RAFH tertile was 96% identical across all graders. Conclusions Our semiautomated image analysis software facilitates stratification of progression risk based on RAFH and enabled a nonexpert grader with minimal training to obtain results comparable to expert graders. Predictions based on RAFH were similar to those of a validated qualitative classification system.

Classical dendritic cells mediate fibrosis directly via the retinoic acid pathway in severe eye allergy

Fibrosis is a shared end-stage pathway to lung, liver, and heart failure. In the ocular mucosa (conjunctiva), fibrosis leads to blindness in trachoma, pemphigoid, and allergy. The indirect fibrogenic role of DCs via T cell activation and inflammatory cell recruitment is well documented. However, here we demonstrate that DCs can directly induce fibrosis. In the mouse model of allergic eye disease (AED), classical CD11b+ DCs in the ocular mucosa showed increased activity of aldehyde dehydrogenase (ALDH), the enzyme required for retinoic acid synthesis. In vitro, CD11b+ DC-derived ALDH was associated with 9-cis-retinoic acid ligation to retinoid x receptor (RXR), which induced conjunctival fibroblast activation. In vivo, stimulating RXR led to rapid onset of ocular mucosal fibrosis, whereas inhibiting ALDH activity in DCs or selectively depleting DCs markedly reduced fibrosis. Collectively, these data reveal a profibrotic ALDH-dependent pathway by DCs and uncover a role for DC retinoid metabolism.

Enhanced visualization of peripheral retinal vasculature with wavefront sensorless adaptive optics optical coherence tomography angiography in diabetic patients

Optical coherence tomography angiography (OCTA) is a promising technique for non-invasive visualization of vessel networks in the human eye. We debut a system capable of acquiring wide field-of-view (>70°) OCT angiograms without mosaicking. Additionally, we report on enhancing the visualization of peripheral microvasculature using wavefront sensorless adaptive optics (WSAO). We employed a fast WSAO algorithm that enabled wavefront correction in <2  s by iterating the mirror shape at the speed of OCT B-scans rather than volumes. Also, we contrasted ∼7° field-of-view OCTA angiograms acquired in the periphery with and without WSAO correction. On average, WSAO improved the sharpness of microvasculature by 65% in healthy eyes and 38% in diseased eyes. Preliminary observations demonstrated that the location of 7° images could be identified directly from the wide field-of-view angiogram. A pilot study on a normal subject and patients with diabetic retinopathy showed the impact of utilizing WSAO for OCTA when visualizing peripheral vasculature pathologies.