Lucentis Approved in China In a recent press release, Novartis announced that it received regulatory approval in China from the State Food and Drug Administration (SFDA) for Lucentis (ranibizumab) to treat wet AMD... FDA to Grant Ocriplasmin Priority Review According to a recent press release issued by ThromboGenics NV, the FDA intended to grant its recently submitted ocriplasmin Biological License Application (BLA) for Priority Review... And More...

Long-Term Follow-Up of Ranibizumab Use for Macular Edema Caused by RVO

To assess long-term safety and efficacy of intraocular ranibizumab injections in patients with macular edema after retinal vein occlusion (RVO), investigators included patients who completed an open-label extension trial of the 12-month Ranibizumab for the Treatment of Macular Edema following Branch Retinal Vein Occlusion: Evaluation of Efficacy and Safety (BRAVO, 304) and Central Retinal Vein Occlusion Study: Evaluation of Efficacy and Safety (CRUISE, 304) trials.

They saw patients at least every 3 months and gave them an intraocular injection of 0.5 mg ranibizumab if they met prespecified retreatment criteria. They noted that primary outcomes were incidence and severity of ocular and nonocular adverse events (AEs). Key efficacy outcomes included mean change from baseline best-corrected visual acuity (BCVA) letter score by Early Treatment Diabetic Retinopathy Study protocol and central foveal thickness.

According to the investigators, in patients who completed month 12, the mean number of injections (excluding month 12 injection) in the sham/0.5-, 0.3/0.5- and 0.5-mg groups was 2.0, 2.4 and 2.1 (branch RVO) and 2.9, 3.8 and 3.5 (central RVO), respectively. They reported that the incidence of study eye ocular serous AEs (SAEs) and SAEs potentially related to systemic vascular endothelial growth factor inhibition across treatment arms was 2% to 9% and 1% to 6%, respectively. The mean change from baseline BCVA letter score at month 12 in branch RVO patients was 0.9 (sham/0.5 mg), –2.3 (0.2/0.5 mg) and –0.7 (0.5 mg), respectively. Moreover, the mean change from baseline BCVA at month 12 in central RVO patients was –4.2 (sham/0.5 mg), –5.2 (0.3/0.5 mg) and –4.1 (0.5 mg), respectively.

The investigators did not identify any new safety events with long-term use of ranibizumab, and they observed that rates of SAEs potentially related to treatment were consistent with prior ranibizumab trials. Reduced follow-up and fewer ranibizumab injections in the second year of treatment were associated with a decline in vision in central RVO patients, but vision in branch RVO patients remained stable. These results suggest that during the second year of ranibizumab treatment of RVO patients, follow-up and injections should be individualized and, on average, central RVO patients may require more frequent follow-up than every 3 months.

Source: Heier JS, Campochiaro PA, Yau L, et al. Ranibizumab for macular edema due to retinal vein occlusions: long-term follow-up in the HORIZON Trial. Ophthalmol. 2012; Feb. 2 [Epub ahead of print]. DOI: 10.1016/j.ophtha.2011.12.005.

Monthly vs. Variable-Dosing of Intravitreal Ranibizumab for the Treatment of CNV Secondary to AMD: Comparison of 1-Year Efficacy

To evaluate the visual acuity results of monthly ranibizumab injections compared with a variable-dosing schedule for the treatment of neovascular age-related macular degeneration (AMD), investigators conducted a retrospective study that compared two cohorts of consecutive patients.

All patients were treatment naïve, with baseline visual acuity of 20/400 or better, and completed 12 months of therapy. In the first group, all patients received monthly injections. In the other group, after 3 monthly loading doses, the investigators used a variable-dosing schedule, based on a monthly clinical assessment and optical coherence tomography.

They included 56 consecutive patients (60 eyes) and reported that at 12 months, the median numbers of injections were 12 and 8, respectively, and the mean change in Snellen visual acuity was an improvement of 0.27 logarithm of the minimum angle of resolution (logMAR) in the monthly treated group versus 0.21 logMAR improvement in the variable-dosing group (p = 0.53). They also noted that in the monthly treated group, 96.8% of eyes lost <0.3 logMARs versus 96.6% of eyes in the variable-dosing group (p = 1.0).

The investigators in this study were able to show that in their clinical setting, patients achieved similar visual acuity results with either monthly injections or with a variable-dosing protocol. Additionally, they observed a trend toward better results with monthly treatment.

Source: Katz G, Giavedoni L, Muni R, et al. Effectiveness at 1 year of monthly versus variable-dosing intravitreal ranibizumab in the treatment of choroidal neovascularization secondary to age-related macular degeneration. Retina. 2011;32(2):293–298.

Ranibizumab in the Treatment of CNV Secondary to AMD

This open-label, multicenter, extension study evaluated the long-term safety and efficacy of multiple intravitreal ranibizumab injections administered at the investigator's discretion in patients with choroidal neovascularization (CNV) secondary to age-related macular degeneration (AMD).

Patients who completed the controlled treatment phase of 1 of 3 prospective, randomized, 2-year clinical trials of ranibizumab were eligible for enrollment and analyses were performed analyses for 3 groups: (1) patients treated with ranibizumab in the initial study (ranibizumab treated-initial; n = 600); (2) patients randomized to control who crossed over to receive ranibizumab (ranibizumab treated-XO; n = 190); and (3) ranibizumab-naïve patients (ranibizumab untreated; n = 63). Ranibizumab 0.5 mg was administered at the investigator's discretion. Adverse events (AEs) and Early Treatment Diabetic Retinopathy Study (ETDRS) best-corrected visual acuity (BCVA) assessments were conducted at study visits every 3 to 6 months. Main outcome measures were incidence and severity of AEs.

There was 1 occurrence of mild endophthalmitis per 3552 HORIZON injections in the ranibizumab treated-initial/ranibizumab treated-XO groups and no serious AE reports of lens damage, retinal tears or rhegmatogenous retinal detachments in the study eyes. It was noted that the proportion of patients with any single post-dose intraocular pressure ≥30 mmHg was 9.2%, 6.6%, and 0%, and the proportion of patients with glaucoma was 3.2%, 4.2%, and 3.2% in the ranibizumab treated-initial, ranibizumab treated-XO, and ranibizumab untreated groups, respectively. Cataract AEs were less frequent in the ranibizumab untreated group: 6.3% versus 12.5% and 12.1% in the ranibizumab treated-initial and ranibizumab treated-XO groups, respectively. It was also reported that the proportion of patients with arterial thromboembolic events as defined by the Antiplatelet Trialists' Collaboration was 5.3% in the ranibizumab treated-initial and ranibizumab treated-XO groups, and 3.2% in the ranibizumab untreated group. At month 48 (2 years of HORIZON), the mean change in BCVA (ETDRS letters) relative to the initial study baseline was 2.0 in the ranibizumab treated-initial group versus –11.8 in the pooled ranibizumab treated-XO and ranibizumab untreated groups.

To conclude, multiple ranibizumab injections were well tolerated for ≥4 years. With less frequent follow-up leading to less treatment, there was an incremental decline of the visual acuity (VA) gains achieved with monthly treatment.

Source: Singer MA, Awh CC, Sadda S, et al. HORIZON: an open-label extension trial of ranibizumab for choroidal neovascularization secondary to age-related macular degeneration. Ophthalmol. 2012; Feb. 6 [Epub ahead of print]. Doi:10.1016/j.ophtha.2011.12.016.

Causes of Unsuccessful Ranibizumab Treatment in Exudative AMD

The authors of the following study aimed to identify the causes of loss of vision after ranibizumab therapy in patients with exudative age-related macular degeneration treated in three clinical settings.

They reported that a retrospective multicentric analysis of 290 consecutive eyes comprising cohorts from 3 clinical settings showed that 21 eyes lost ≥15 letters on the Early Treatment Diabetic Retinopathy Study chart 1 year after the start of ranibizumab treatment. They also indicated that fundus images of these eyes were analyzed by two independent readers to investigate the causes of visual loss. The study authors compared the three cohorts and a search was made for factors predisposing to visual loss. They then performed a second analysis to compare the baseline characteristics of patients who gained (visual acuity gainers) or lost (visual acuity losers) ≥15 letters.

According to the authors, among the 290 eyes included, the proportions from each center experiencing visual loss were not significantly different (mean, 7.24%, p = 0.2631). Mean visual loss of affected eyes was 27 letters and they found no significant difference between these eyes and others with regard to age and gender of patients, laterality, type of choroidal neovascularization, number of visits, or initial visual acuity. They noted that visual loss was secondary to the progression of atrophy in eight eyes, fibrosis in five eyes, a combination of fibrosis and atrophy in three eyes, severe subretinal hemorrhage in three eyes, and retinal pigment epithelial tear in two eyes. Moreover, they observed significant difference between visual acuity gainers and losers for 2 parameters: age of patients, 80.9 ± 5.3 years in visual acuity losers versus 77.5 ± 7.3 years in visual acuity gainers (p = 0.0473) and visual acuity at diagnosis, respectively, 56.2 ± 11.2 versus 49.0 ± 12.0 (p = 0.0288).

To conclude, although uncommon, visual loss may occur during ranibizumab treatment and is because of the natural course of age-related macular degeneration in most cases.

Source: Cohen SY, Oubraham H, Uzzan J, et al. Causes of unsuccessful ranibizumab treatment in exudative age-related macular degeneration in clinical settings. Retina. 2012; Jan. 17 [Epub ahead of print]. DOI: 10.1097/IAE.0b013e318240a516.

Changes in Microperimetry Following Intravitreal Bevacizumab Injection for Exudative AMD

Researchers in Turkey evaluated the effect of intravitreal bevacizumab on macular function in the cases of exudative age-related macular degeneration (AMD). They found that intravitreal bevacizumab therapy induced a significant increase in mean retinal sensitivity and significant decrease in mean absolute scotoma size during 6 months. They also determined that the MP-1 microperimetry proved to be a valuable tool in the evaluation of functional benefit of exudative AMD therapy with intravitreal bevacizumab.

In this study, they included 21 eyes of 21 patients with exudative AMD and in each eye, at baseline and 1, 3 and 6 months after intravitreal bevacizumab injection, they assessed logMAR visual acuity, central 4° macular sensitivity, absolute scotoma size, fixation stability and fixation location by MP-1 microperimetry and optical coherence tomography (OCT) foveal morphologic changes. Following the initial treatment phase, which included three consecutive injections, they based their decision to re-treat on OCT and clinical findings. Subsequent injections could be administered at least 1 month after the previous injection period according to the OCT-guided treatment regimen.

The researchers in this study found that mean retinal sensitivity within central 4° (12 points) area had increased from 3.69 ± 3.44 dB at baseline to 7.16 ± 3.27 dB at month 6. In all controls after the treatment, they reported a significant increase in logMAR visual acuity (p<0.001) and MP-1 retinal sensitivity (p<0.001). Mean absolute scotoma in test point location had decreased significantly from 12 of the 76 applied test point locations measured at baseline to five test point locations (–7 test point locations; p<0.001) at month 6 showing statistical significance. Fixation properties had preserved in all patients 6 months after intravitreal bevacizumab treatment.

Source: Ozdemir H, Karacorlu M, Sentur F, et al. Microperimetric changes after intravitreal bevacizumab injection for exudative age-related macular degeneration. Acta Ophthalmol. 2012;90(1):71–75.

Relationship Between Exudative AMD and Polymorphisms in C2, CFB and C3 in a Korean Population

This study investigated the association of genetic polymorphisms in complement component 2 (C2), complement factor B (CFB) and complement component 3 (C3) with exudative age-related macular degeneration (AMD) in a Korean population and the gene–gene and gene–environment interactions in the development of AMD.

A total of six SNPs that are located in the C2 (rs547154, rs9332739), CFB (rs4151667, rs641153) and C3 (rs1047286, rs2230199) genes were genotyped in 350 samples comprised of 153 cases, 197 controls. The risk allele frequencies for rs547154 in C2 were 6.54% and 8.12% in AMD patients and controls, while those for rs641153 in CFB were 6.54% and 8.63% in AMD patients and controls. The risk allele frequency for rs9332739 in C2 (AMD, 0.65%, control, 2.03%) and rs4151667 in CFB (AMD, 0.65%, control, 1.78%) was very low.

It was reported that the protective allele of four SNPs was not significantly associated with decreased risk for AMD (p = 0.427, p = 0.199, p = 0.312, p = 0.303, respectively) and that the homozygotes for the protective allele of four SNPs were not significantly associated with decreased risk for AMD (p = 0.324, p = 0.474, p = 0.309, p = 0.411, respectively). Furthermore, the genetic effect of two SNPs in C3 could not be investigated because the variants were not observed. There was no evidence to support an interaction of these SNPs with LOC387715/HTRA1 variants or with environmental exposure like smoking.

In conclusion, the genetic effect of C2, CFB and C3 polymorphisms, which are known to be important for AMD in Caucasian, were not significant in the Korean population. The low minor allele frequency of these SNPs in Koreans might have affected the results of this study. It is suggested that ethnic differences in the roles of C2, CFB and C3 in conferring a risk of AMD be further investigated.

Source: Kim SJ, Lee SJ, Kim NR, Chin HS. Association of polymorphisms in C2, CFB and C3 with exudative age-related macular degeneration in a Korean population. Exp Eye Res. 2012; Jan. 18 [Epub ahead of print]. DOI: 10.1016/j.exer.2012.01.005.

Influence of CFH, VEGF and HTRA1 Promoter Genotype on the Response to Intravitreal Ranibizumab Therapy for Neovascular AMD

Investigators in the United Kingdom sought to examine an association between genotype for three single nucleotide polymorphisms strongly associated with the development of age-related macular degeneration (AMD) and the early response to treatment with intravitreal ranibizumab for neovascular AMD.

They recorded best-corrected visual acuity (BCVA) letter score at baseline and at each subsequent visit, as well as age, sex, smoking history, lesion type and the number of injections. They also obtained genotypes for rs11200638 in HTRA1, rs1061170 in CFH and rs1413711 in VEGF and analyzed data with treatment response at month 6 as both a binary (>5 letter improvement vs. ≤5 letter gain) and a linear trait.

This initial cohort consisted of 104 Caucasian neovascular AMD patients treated with intravitreal ranibizumab. The study investigators observed trends toward a more favorable outcome with the higher AMD risk genotypes in CFH and VEGF in both the linear and binary models and in HTRA1 in the linear model alone. For CFH, they noted that mean letter score change after 6 months was +1.6, +5.9 and +7.2 letters for the TT, TC and CC genotypes and they saw a >5 letter gain in 34.6%, 56.6% and 56%, respectively. For VEGF, mean letter score change after 6 months was +1.3, +5.8 and +7.4 letters for the TT, TC and CC genotypes and they noted a >5 letter gain in 40%, 55.8% and 51.9%, respectively. Moreover, for HTRA1 in, mean letter score change was +2.2, +7.5 and +2.9 letters for the GG, GA and AA genotypes.

This study reports preliminary evidence suggesting that the higher AMD risk genotypes in CFH, VEGF and HTRA1 may influence the short-term response to treatment with ranibizumab for neovascular AMD.

Source: McKibbin M, Ali M, Bansal S, et al. CFH, VEGF and HTRA1 promoter genotype may influence the response to intravitreal ranibizumab therapy for neovascular age-related macular degeneration. Br J Ophthalmol. 2012;96(2):208–212.

Treatment of Subfoveal Idiopathic CNV with Intravitreal Bevacizumab

To evaluate the visual and anatomic outcomes of intravitreal bevacizumab in patients with subfoveal idiopathic choroidal neovascularization (CNV), the authors of this Chinese study conducted a prospective, nonrandomized, interventional case series.

They included 40 patients with subfoveal idiopathic CNV, whose eyes they treated with a single intravitreal injection of 1.25 mg bevacizumab followed by as-needed dosing indicated by the presence and recurrence of intraretinal edema, subretinal fluid (SRF) or pigment epithelial detachment (PED), based on optical coherence tomography (OCT) performed monthly. They authors observed visual, clinical, angiographic and anatomic changes over a 12-month follow-up period.

They found that after 12 months of follow-up, the mean logarithm of minimal angle of resolution (logMAR) best-corrected visual acuity (BCVA) improved from 0.53 (20/68 in Snellen equivalent) at baseline to 0.29 (20/39 in Snellen equivalents; p<.001). They also noted that mean central retinal thickness determined by OCT decreased from 321 µm to 237 µm (p<.001) and that all eyes (100%) had stable or improved vision, and 28 (70%) showed an improvement of 2 lines or more. According to the study authors, all lesions were in the cicatricial stage of CNV at 12 months of follow-up, with no leakage of fluorescein in the late phase of fluorescein angiography and no intraretinal edema, SRF and/or PED detected by OCT. Furthermore, they observed no drug-related systemic or ocular side effects.

In conclusion, intravitreal bevacizumab is generally well tolerated and improves BCVA in eyes with subfoveal idiopathic CNV over a period of 12 months. Large, randomized, controlled, long-term clinical trials are required to further evaluate the efficacy and optimal strategy of this treatment modality.

Source: Zhang H, Liu ZL, Sun P, Gu F. Intravitreal bevacizumab for treatment of subfoveal idiopathic choroidal neovascularization: results of a 1-year prospective trial. Am J Ophthalmol. 2012;153(2):300–306.

Use of Near-Infrared AF in Patients with Idiopathic CNV

The authors of this Japanese retrospective observational consecutive case series investigated near-infrared autofluorescence (IR-AF) patterns and related changes in patients with idiopathic choroidal neovascularization (CNV) treated with intravitreal bevacizumab (IVB).

They intravitreally injected bevacizumab into 12 eyes of 12 patients with idiopathic CNV as the primary treatment. They also performed color fundus photographs, optical coherence tomography (OCT), fluorescein angiography and indocyanine green angiography (ICGA) and autofluorescence imaging short-wavelength and near-infrared autofluorescence (SW-AF and IR-AF) at baseline. Finally, the authors evaluated changes in the autofluorescence patterns after IVB.

They reported that all 12 eyes had classic CNV on fluorescein angiography at baseline and that OCT showed CNV above the retinal pigment epithelium (RPE) in all eyes. Following treatment, they noted that final best-corrected visual acuity improved significantly (p<.001) compared with baseline. According to the study authors, IR-AF showed ring-shaped hyperautofluorescence surrounding the CNV corresponding to the dark rim on ICGA in 6 of the 12 eyes on IR-AF at baseline. They also found that during the follow-up period after IVB, all 12 eyes had ring-shaped hyperautofluorescence. Additionally, they observed that the intensity of the ring-shaped autofluorescence and its contrast increased as the CNV regressed. Moreover, the contrast of the ring-shaped autofluorescence partially decreased in all 3 eyes with a recurrence.

Ring-shaped hyperautofluorescence on IR-AF in the eyes with idiopathic CNV may indicate an involutional process of CNV enveloped by the RPE because its area corresponded to the dark rim on ICGA that reflects regression of idiopathic CNV. IR-AF can be a useful noninvasive adjunctive examination to evaluate the involution of CNV.

Source: Toju R, Iida T, Sekiryu T, et al. Near-infrared autofluorescence in patients with idiopathic submacular choroidal neovascularization. Am J Ophthalmol. 2012;153(2):314–319.

Factors Linked to Enlargement of Chorioretinal Atrophy Following Treatment with Intravitreal Bevacizumab for mCNV

Japanese researchers reviewed the medical charts of 27 eyes with a myopic choroidal neovascularization (mCNV) that had received intravitreal bevacizumab to determine the factors significantly associated with an enlargement of the area of a chorioretinal atrophy (ChRA) after intravitreal bevacizumab to treat mCNV.

The ophthalmic examinations included measurements of the best-corrected visual acuity, visual fields with the Humphrey 10-2 Field Analyzer (FA), fluorescein angiography and indocyanine green angiography. Additionally, the area of the mCNV and the ChRA were measured on the FA images.

The researchers reported that eyes with an enlargement of the ChRA had significantly larger mCNVs at the baseline, a greater reduction in the size of the mCNV, a higher incidence of subretinal hemorrhage, longer duration of follow-up, received more injections of intravitreal bevacizumab and had a greater decrease of retinal sensitivity (p≤0.041). They also noted that multiple regression analyses showed that the factors most significantly associated with an enlargement of the ChRA were the CNV size at baseline, the number of intravitreal bevacizumab injections and the duration of the follow-up period (p<0.0001).

Their findings showed that eyes with a larger CNV at the baseline and longer follow-up period had a greater risk of developing a ChRA like non-treatment, even if IVB treatment was performed for mCNV.

Source: Uemoto R, Nakasato-Sonn H, Kawagoe T, et al. Factors associated with enlargement of chorioretinal atrophy after intravitreal bevacizumab for myopic choroidal neovascularization. Graefes Arch Clin Exp Ophthalmol. 2012; Jan. 26 [Epub ahead of print]. DOI: 10.1007/s00417-011-1921-4.

RNFL Thickness in RP Evaluated Using SD-OCT

Scientists conducted the following study with the goal of measuring the peripapillary retinal nerve fiber layer (RNFL) thickness using spectral-domain optical coherence tomography (SD-OCT) in patients with retinitis pigmentosa (RP).

They performed a complete ocular examination, including best-corrected visual acuity (BCVA) using a Snellen chart, slit lamp biomicroscopic examination and Goldmann applanation intraocular pressure (IOP) measurement on 50 eyes of 30 patients with RP as well as a dilated fundus examination using both direct and indirect ophthalmoscopy. They also used an OPKO SD-OCT (OPKO Instrumentations) to measure peripapillary RNFL thickness.

Based on the scientists' report, the mean (±SD) age of the study cohort was 45.8 (±16.3) years. They also noted that of the 50 eyes, 18 (36%) showed a thinning of the peripapillary RNFL in 1 or more quadrants and 21 (42%) showed a thickening of the peripapillary RNFL in 1 or more quadrants. Additionally, four eyes (8%) showed both thinning and thickening of the peripapillary RNFL thickness. The overall circumferential RNFL thickness of the 14 eyes that showed only thinning in at least 1 quadrant was 78.78 µm, the scientists observed. They also found that, for the 17 eyes that showed only thickening in at least 1 quadrant, the RNFL thickness was 119.69 µm and that the values of the eyes with thinning and the eyes with thickening were significantly different from normal (t = 6.31 and p<0.01 for thickening; t = 3.62 and p<0.01 for thinning).

Using SD-OCT testing, the study scientists demonstrated in the current study that the peripapillary RNFL thickness in patients with RP can be decreased, increased or maintained within normal limits. Assessment of the RNFL thickness seems prudent in these patients, particularly for identifying notable degrees of RNFL thinning in those being considered for future therapeutic trials.

Source: Anastasakis A, Genead MA, Mcanany JJ, Fishman GA. Evaluation of retinal nerve fiber layer thickness in patients with retinitis pigmentosa using spectral-domain optical coherence tomography. Retina. 2012;32(2):358–363.

Assessment of Physiological Variation of RNFL Thickness and Macular Volume in Humans

With the introduction of spectral domain optical coherence tomography (SD-OCT), changes in the retinal nerve fiber layer (RNFL) thickness and macular volume (MV) can be detected with a high precision. Scientists sought to determine whether there is a physiological quantifiable degree of variation of these structures in humans.

They assessed weight, height, hydration status, RNFL thickness (ring scan, 12 degrees around the optic nerve head) and MV (20x20 degrees) in 69 subjects (44 runners, 25 controls) using the Heidelberg Spectralis with eye-tracking function during a 10 km charity run at VU University Amsterdam. They assessed the SD-OCT scans before running (normal status), after running (more dehydrated status) and 1–1.5 hours after finishing the run (rehydrated status) and measured controls (who did not participate in the running event) at the same time intervals as the runners. The scientists assessed changes over time by general linear models (GLM), correcting for repeated measurements.

In runners, they observed a significant increase of both the RNFL thickness (94.4 µm (baseline) to 95.2 µm (rehydration), p=0.04) and MV (288.9 µm (baseline) to 291.0 µm (rehydration), p<0.001) over time. They also found that controls did not show significant changes over time and that, anatomically, the physiological change of RNFL thickness was most marked in the nasal sectors.

To conclude, this prospective study demonstrated a significant physiological variation of the RNFL thickness and MV at a proportion which, on an individual patient level, may be relevant for longitudinal studies in neurodegenerative diseases.

Source: Balk LJ, Sonder JS, Strijbis EM, et al. The physiological variation of the retinal nerve fibre layer thickness and macular volume in humans as assessed by spectral-domain optical coherence tomography. Invest Ophthalmol Vis Sci. 2012; Jan. 20 [Epub ahead of print]. DOI: 10.1167/iovs.11-8209.

A Look at Automated Multi-Level Pathology Identification Techniques for Abnormal Retinal Images

The aim of the following study in India was to automatically classify abnormal retinal images from four different categories using artificial neural networks with a high degree of accuracy in minimal time to assist the ophthalmologist in subsequent treatment planning.

A total of 420 abnormal retinal images from four different categories (non-proliferative diabetic retinopathy, central retinal vein occlusion, central serous retinopathy and central neo-vascularization membrane) were used. Furthermore, green channel extraction, histogram equalization and median filtering were used as image pre-processing techniques, followed by texture-based feature extraction. The application of Kohonen neural networks for pathology identification was also explored.

The described approach yielded an average classification accuracy of 97.7% with ± 0.8% deviation for individual categories. It was noted that the average sensitivity and the specificity values were 96% and 98%, respectively. The time taken by the Kohonen neural network to achieve these accurate results was 300 ± 40 s for the 420 images.

This study suggests that the approach described can act as a diagnostic tool for retinal disease identification. It was determined that simultaneous multi-level classification of abnormal images is possible with high accuracy using artificial neural networks. The results also suggest that the approach is time-efficient, which is essential for ophthalmologic applications.

Source: Anitha J, Vijila CK, Selvakumar AI, et al. Automated multi-level pathology identification techniques for abnormal retinal images using artificial neural networks. Br J Ophthalmol. 2012;96(2):220–223.