Ophthalmic Surgery, Lasers and Imaging Retina

CLINICAL SCIENCE 

Relation Between Retinal Thickening and Clinically Visible Fundus Pathologies in Mild Nonproliferative Diabetic Retinopathy

Hirokazu Nishiwaki, MD; Mehnaz Shahidi, PhD; Susan Vitale, PhD, MHS; Judith Alexander, BA; Sanjay Asrani, MD; Marek Mori, MS; Norman P Blair, MD; Ran Zeimer, PhD

Abstract

* BACKGROUND AND OBJECTIVE: To determine the association of retinal thickening (RT) with clinically observable retinal pathologies in eyes with mild nonproliferative diabetic retinopathy.

* PATIENTS AND METHODS: Using an objective quantitative imaging method (Retinal Thickness Analyzer), the ratio relative to normal RT (RTI) was measured in 23 eyes with and 35 eyes without clinically observable diabetic rundus pathology. RTI was analyzed in relation to presence of mild diabetic retinal lesions in the ±0.5 mm vicinity.

* RESULTS: The percent of eyes with RTI significantly above normal values did not differ significantly between eyes with and without retínopathy (30% vs 34%). Mean RTI was not associated with local presence of microaneurysms (P=0.92), soft exudates (P=Q. 55), or retinal hemorrhages (P=0.31). Areas without hard exudates had significantly greater mean RTI (1.10) than areas with exudates (0.97, P=0.009).

* CONCLUSION: In diabetic patients with mild retinopathy, areas with and without clinically observable retinal pathologies had similar retinal thickness. We conclude that clinical strategies for detection of retinal thickening should not be limited to areas with visible fundus pathologies.

[Ophthalmic Surg Lasers 2002;33:127-134.]

Abstract

* BACKGROUND AND OBJECTIVE: To determine the association of retinal thickening (RT) with clinically observable retinal pathologies in eyes with mild nonproliferative diabetic retinopathy.

* PATIENTS AND METHODS: Using an objective quantitative imaging method (Retinal Thickness Analyzer), the ratio relative to normal RT (RTI) was measured in 23 eyes with and 35 eyes without clinically observable diabetic rundus pathology. RTI was analyzed in relation to presence of mild diabetic retinal lesions in the ±0.5 mm vicinity.

* RESULTS: The percent of eyes with RTI significantly above normal values did not differ significantly between eyes with and without retínopathy (30% vs 34%). Mean RTI was not associated with local presence of microaneurysms (P=0.92), soft exudates (P=Q. 55), or retinal hemorrhages (P=0.31). Areas without hard exudates had significantly greater mean RTI (1.10) than areas with exudates (0.97, P=0.009).

* CONCLUSION: In diabetic patients with mild retinopathy, areas with and without clinically observable retinal pathologies had similar retinal thickness. We conclude that clinical strategies for detection of retinal thickening should not be limited to areas with visible fundus pathologies.

[Ophthalmic Surg Lasers 2002;33:127-134.]

INTRODUCTION

Diabetic macular edema is one of the leading causes of visual impairment among adults in the United States.1 This disorder is known to be associated with a worsening of glycémie control, increased duration of diabetes, and increased biood pressure.2"7 Since diabetic macular edema can be difficult to detect, it is most sensitively identified with contact lens biomicroscopy by a retina specialist.8 Unfortunately, an expert examination is subjective and difficult to document. For this reason, in most major clinical trials using macular edema as a study endpoint, expert grading of stereoscopic fundus photographs has been used.9'10 However, a recent study by Yang et al" demonstrated that an expert grader often (in 30% of cases) does not recognize retinal thickening that can be definitively detected by a reliable, objective measure of retinal thickness. Additionally, the results of this study indicated that areas with thickening were more likely to be identified in the proximity of hard exudates.

The Early Treatment Diabetic Retinopathy Study Research Group (ETDRS)12 demonstrated that focal laser treatment for clinically significant macular edema can reduce the risk of a 3-line loss of vision (ie, doubling of the visual angle) by 50% over a 3-year period. Based on these results, current clinical practice patterns recommend laser treatment for patients with CSME. "The precise method for treatment13'15 consists of applying focal laser burns to microaneurysms within a thickened retina, and of applying a light grid of laser burns to thickened areas with dimise leakage (as identified by fluorescein angiogram). In the absence of retinal pathology, the retina is transparent and perception of stereopsis is limited, making visualization difficult for even an expert observer.16 However, when microaneurysms, hard exudates, or retinal hemorrhages are present, their location in the inner retinal layers forms a basis for stereoscopic perception. In the present study, the association between visible fundus pathology and retinal thickening was studied in diabetic eyes with early nonproliferarive retinopathy.

Table

Table 1. Characteristics of Patients*

Table 1. Characteristics of Patients*

PATIENTS AND METHODS

Patients

Patients with mild nonproliferative diabetic retinopathy were recruited from the University of Illinois at Chicago. One eye from each patient that met the following requirements was included in the study: clear media, minimum of 5 mm dilated pupil, no history of laser treatment, and an absence of proliferative retinopathy. The study protocol was reviewed and approved by the Institutional Review Board of the University of Illinois at Chicago. AJl patients gave informed consent to participation in the study. A total of 58 diabetic patients were recruited, 35 eyes without and 23 eyes with retinopathy (presence of visible fundus pathologies in the posterior pole). Characteristics of the eyes are shown in Table 1. Mean age and duration of diabetes was similar for patients with and without retinopathy.

Retinal Thickness Analysis

The principle of retinal thickness analysis has been previously described.17'18 Briefly, it is based on projecting a thin laser slit obliquely on the retina and viewing it at an angle in a manner similar to slit-lamp biomicroscopy. The separation between the reflections (and scatter) of the laser beam from the vitreoretínal and the chorioretinal interraces provides a measure of the retinal thickness. The results presented here were obtained with our prototype of the nonscanning version of the Retinal Thickness Analyzer (RTA).16 A green (540 mm) HeNe laser was focused into a slit 1 5 microns in width and 2 mm in length on the retina.

Figure 1. Areas of the posterior pole measured with the RTA. Pathology was recorded for the ± 0.5 mm area (represented by the rectangular box) surrounding each 2 mm linear scan (vertical line centered in each rectangular box), and also in the entire posterior pole.

Figure 1. Areas of the posterior pole measured with the RTA. Pathology was recorded for the ± 0.5 mm area (represented by the rectangular box) surrounding each 2 mm linear scan (vertical line centered in each rectangular box), and also in the entire posterior pole.

Prior to imaging, the pupil was dilated with 1% tropicamide and a water-filled contact lens was placed on the eye. The patient was asked to view an internal fixation target and to fixate at each one of 10 arranged internal targets (Figure 1). A magnified fundus image showing the intersection of the laser slit with the retinal interfaces was captured simultaneously with the acquisition of the laser slit scan by a 35 mm camera attached to the slit lamp. This image of the fundus surrounding the laser slit was used to document the exact location of the retinal thickness measurement. For each of the areas of interest (Figure 1), a 2 mm linear scan was obtained.

The images were analyzed by an automated, operator free, upgraded software algorithm.19 Along the 2 mm scan, the locations of the 2 retinal interfaces were detected. This was achieved by fitting a Lorenztian curve to the predominant peak to determine the location of one interface. This fitted curve was then subtracted from the profile, yielding a second peak that corresponded to the second interface. The separation between the location of the 2 peaks was determined and converted to retinal thickness units. The operator-free algorithm performed automatic quality tests and deleted measured points along the 2 mm scan that deviated from their neighbors by more than a preset value. If more than 1 scan was acquired for a given location, the algorithm identified the scan with the better image quality. No scan passed directly through a retinal lesion. (It is possible that the presence of hard exúdate or hemorrhage could affect the retinal thickness measurement in that specific location; however, this situation did not occur in these data.) The retinal thickness values over each I mm segment were then averaged to provide retinal thickness measurements at 20 locations in each eye (2 adjacent measurements at each of the 10 locations). We did not use any individual correction for refractive error in this study because the use of the contact lens minimized the influence of the optics of the eye.

Retinal thickness for each location was normalized by the corresponding average thickness in normal eyes20 to provide a retinal thickness index (RTI). RTI values of 1.16 or higher were outside the 95% confidence interval for normal thickness and thus considered outside normal limits. From the data obtained in each eye, the foveal RTI (thickness at the fovea! location) and maximum RTI (maximum thickness at any measured location) was also determined. Maximum RTI may be more useful than mean RTI in identifying local areas of retinal edema.

Fundus Photography Grading

Color fundus photography of a 30° field, centered on the macula in each eye, was performed by a trained ophthalmic photographer. The locations of the retinal thickness scans for each eye were marked on a translucent overlay on the fundus photographs. The fundus photographs were then graded for presence and type of retinal pathologies by an expert reader (JA). The presence and number of microaneurysms, retinal hemorrhages, soft exudates, and hard exudates were recorded within 0.5 mm of the location of the retinal thickness measurements (Figure 1), and also for the entire area covered by the fundus photograph.

Statistical Analysis

Descriptive analyses were performed at the eye level. Wilcoxon rank-sum tests were used to test whether median RTI differed between subgroups of eyes. Analyses were performed on all eyes and the subgroup of eyes with retinopathy. To examine the association between RTI and presence of pathology, Generalized Estimating Equation (GEE) models were used to account for the correlations of areas within eyes.21 The dependent (y) variable was the RTI. The independent (x) variable was either presence/absence of a particular type of pathology or the number of pathologies of that type. A GEE model was calculated separately for each pathology type. The estimate for beta and its robust standard error estimate were used to test for statistical significance, using the WaId chisquared statistic. P-values of 0.05 or less were considered statistically significant. Separate models for RTI in the foveal area, and for maximum RTI, were also examined.

Table

Table 2. Association of RTI with Presence/Absence of Retinal Pathology Within 0.5 mm, Based on Areas Within Eyes (GEE Model to Account for Correlation of Areas Within the Same Eye)

Table 2. Association of RTI with Presence/Absence of Retinal Pathology Within 0.5 mm, Based on Areas Within Eyes (GEE Model to Account for Correlation of Areas Within the Same Eye)

RESULTS

Areas with and without RTI information did not differ with respect to presence or number of microaneurysms, retinal hemorrhages, soft exudates, or hard exudates (P=GJ, 0.6, 0.2, and 0.7; P=0.6, 0.4, 0.2, and 0.5, respectively).

Mean RTI (over all areas and all eyes) was similar in eyes with and without retinopathy 1.097 (range, 0.89-1.64, N=23) and 1.098 (range, 0.87-1.36, N=35), respectively. In 23 (66%) of 35 eyes without retinopathy, and in 16 (70%) of 23 eyes with retinopathy, the mean RTI was within normal limits (RTI =£1.16). Overall, eyes with retinopathy had little pathology in the ± 0.5 mm vicinity of the retinal thickness measurements. The most frequently observed pathology in these areas was retinal hemorrhages.

Maximum RTI (over all areas) was similar in eyes with and without retinopathy 1.34 (range, 1.02 - 2.20, N=23) and 1.30 (range, 1.05-1.59, N=35), respectively. In 10 (29%) of 35 eyes without retinopathy, and in 8 (35%) of 23 eyes with retinopathy, the maximum RTI was within normal limits (RTI =£1.16). The maximum RTI, averaged over eyes in which no fundus pathology was present at the sites measured, was compared to that of the eyes without retinopathy. The average of the maximum RTI for eyes with microaneurysms, hard exudates, retinal hemorrhages, soft exudates, or any pathology, was 1.37, 1.35, 1.36, 1.35, and 1.45, respectively: none of these values differed significantly from the average of the maximum RTI for eyes without retinopathy (1.30, P>0.6) for all comparisons.

Local Association of Pathology With RTI

To determine the local association of visible fundus pathologies with RT, the pathology gradings within ± 0.5 mm areas surrounding each RTA linear scan were used and GEE models were employed to adjust for the correlation of areas within an eye. Median number of areas with data was 12 for eyes without retinopathy and 1 1 for eyes with retinopathy (range: 6 to 1 0 and 6 to 20, respectively; mean: 12.2 and 1 1.7, respectively).

Results of analyses for the presence/absence of each type of pathology are shown in Table 2, The presence of microaneurysms and retinal hemorrhages was not associated with retinal thickness. Areas with soft exudates tended to have greater RTIs, but this relation was not statistically significant. Areas with hard exudates had, on average, RTIs that were 0.13 less than areas without hard exudates (P= 0.009) (mean RTI for areas with hard exudates: 0.97; mean RTI for areas without hard exudates: 1.10). Similar results were obtained by analyzing the data according to the number of pathologies within 0.5 mm of the retinal thickness measurement locations (Table 3) . For each additional hard exúdate, RTI decreased by 0.06, on average (P=.0004).

Table

Table 3. Association of RTI With Number of Retinal Pathologies Within 0.5 mm, Based on Areas Within Eyes (GEE Model to Account for Correlation of Areas Within the Same Eye)

Table 3. Association of RTI With Number of Retinal Pathologies Within 0.5 mm, Based on Areas Within Eyes (GEE Model to Account for Correlation of Areas Within the Same Eye)

Maximum RTI and Total Number of Pathologies in the Posterior Pole

The association of the total number of pathologies in the posterior pole with the maximum measured RTI for each eye was determined for each type of retinal lesion by using GEE models (Table 4). The total number of microaneurysms and retinal hemorrhages in the posterior pole was not associated with the maximum RTI. The maximum RTI was slighdy greater in eyes with more hard exudates and slightly less in eyes with less soft exudates, but the differences were not statistically significant.

Table

Table 4. Association of Maximum RTI (of All Measured Areas Within an Eye) With Total Number of Retinal Pathologies in Posterior Pole, Based on Areas Within Eyes (GEE Model to Account for Correlation of Areas Within the Same Eye)

Table 4. Association of Maximum RTI (of All Measured Areas Within an Eye) With Total Number of Retinal Pathologies in Posterior Pole, Based on Areas Within Eyes (GEE Model to Account for Correlation of Areas Within the Same Eye)

Fovea! RTI and Associated Pathology

There was no statistically significant difference in foveal retinal thickness between eyes with and without retinopathy: mean retinal thickness in the foveal area was 1.14 and 1.06 for eyes with and without retinopathy, respectively (% with normal foveal retinal thickness was 74% for no retinopathy [25/34] and 70% for eyes with retinopathy [14/2O]). In the foveal area, only 7 eyes had some visible pathology: 3 eyes had microaneurysms, 2 eyes had retinal hemorrhages, and 2 eyes had soft exudates. The association between retinal pathology and foveal RTI was examined (Table 5). The results indicated a lack of a statistically significant relationship between foveal thickness and both the presence and number of retinal pathologies.

Table

Table 5. Association of Foveal Mean RTI With Presence/Absence of Retina! Pathology, Eyes With at Least Some Retinopathy Anywhere in the Eye (GEE Model to Account for Correlation of Areas Within the Same Eye)

Table 5. Association of Foveal Mean RTI With Presence/Absence of Retina! Pathology, Eyes With at Least Some Retinopathy Anywhere in the Eye (GEE Model to Account for Correlation of Areas Within the Same Eye)

DISCUSSION

Detection of diabetic macular edema is necessary to initiate treatment and prevent vision loss. It has been known for some time that clinical observation using contact lens biomicroscopy is a more sensitive method of detecting retinal thickening than grading of stereoscopic fundus photographs.8 However, clinical observation depends on experience, skill, and the ability to perceive stereoscopic depth. In a normal retina, the tissue is transparent and an impression of retinal thickness can be difficult to obtain.22 When exudates, hemorrhages, or microaneurysms are present within the layers of the retina, a basis for stereoscopic perception is provided. This is a possible explanation of our previous finding that retinal thickening was more likely to be detected by stereophotography grading when hard exudates were present.23

A method for identifying retinal thickening in a reliable and accurate way could be useful in assessing the need for treatment and in following the results of treatment. Previous studies24 showed that a decrease in retinal thickness after treatment is more likely with a lower degree of pretreatment thickening. Accurate and early detection of retinal thickening may therefore enhance the effectiveness of laser treatment.

In the present study, we investigated the relationship between the presence of different types of visible fundus pathologies and retinal thickness in diabetic patients with mild nonproliferative retinopathy. Neither the presence nor number of microaneurysms, retinal hemorrhages, or soft exudates was associated with alterations in the retinal thickness. These findings are in agreement with those of Schaudtg et al,25 who found no significant differences in retinal thickness measured by optical coherence tomography between persons with and without diabetic retinopathy (none had clinically significant macular edema or proliferative retinopathy), except for the inferotemporal quadrant measurement, in which those with retinopathy had, on average, retinal thickness 1 5 µ?? greater than those without. In our current report, the group with mild retinopathy had a presence and number of hard exudates directly associated with decreased thickness in the adjacent retinal area; areas with no hard exudates tended to be thicker than areas with hard exudates, although most values were within normal limits for both groups. A possible explanation of this finding is, as edema develops, fluid, proteins, and lipids leak into the intraretinal (extracellular) space. As edema resolves, the fluid would be reabsorbed more quickly than the lipids, which may remain as markers of previous edema. However, this speculation needs to be confirmed in further studies. This result cannot be attributed to changes in the reflection or scattering of the laser light because of the presence of the hard exudates. First, the thickness measurements were not obtained directly on the hard exudates; and second, increased scattering from the hard exudates would broaden the laser width and thus provide larger values for thickness, rather than smaller, contrary to our findings. Current thinking among vitreoretinal specialists is that there is a correlation between hard exudates, especially multiple fine hard exudates, and hard exúdate rings with retinal thickening in moderate to severe diabetic retinopathy. Since our findings were observed in eyes with only mild nonproliferative diabetic retinopathy (with an average of 2,2 hard exudates in the posterior pole), additional studies are necessary to determine the relationship in more advanced retinopathy.

The average retinal thickness, for the fovea and all other locations, was often within the normal range in eyes with and without visible fundus pathologies. In eyes with visible fundus pathologies such as mlcroaneurysms, retinal hemorrhages, soft exudates, and hard exudates, the maximum thickness was outside the normal range and slightly higher (not statistically significantly so) than in eyes without pathologies. Average retinal thickness may not be the optimal measure for detecting retinal edema because the averaging process will tend to mask local areas of thickening. Maximum thickness may be a better measure for the purpose of detecting localized thickening. Even so, in these eyes with mild nonproliferative diabetic retinopathy, the maximum RTI did not differ significantly between areas with and without visible fundus pathologies.

A possible limitation of this study is the infrequent occurrence of retinal pathology within 0.5 mm of the thickness measurements. The power to detect significant thickness differences between eyes with and without local pathology is diminished by the small comparison group. However, the differences in mean thickness between areas with and without local pathology were negligible, regardless of the level of statistical significance. We contend that clinical strategies for detection of retinal thickening should not be limited to areas with visible fundus pathology because these data show that visible fundus pathology is not associated with local retinal thickening.

REFERENCES

1. National Diabetes Data Group, National Institute of Diabetes and Digestive and Kidney Diseases. Diabetes in America. 2nd ed. Bethesda, MD: National Institutes of Health.

2. Klein R, Moss SE, Klein BE, Davis MD, DeMets DL. The Wisconsin Epidemiologie Study of Diabetic Retinopathy. XI. The incidence of macular edema. Ophthalmology. 1989;96:1501-1510.

3. Klein R, Klein BE, Moss SE, Cruickshanks KJ. The Wisconsin Epidemiologie Study of Diabetic Retinopathy. XV. The long-term incidence of macular edema. Ophthalmology. 1995:102:7-16.

4. Vitale S, Maguire MG, Murphy RP, et al. Clinically significant macular edema in type I diabetes. Incidence and risk factors. Ophthalmology. 1995U02-.1170-1176.

5. Klein R, Klein BE, Moss SE, Cruickshanks KJ. The Wisconsin Epidemiologie Study of Diabetic Retinopathy: XVII. The 14-year incidence and progression of diabetic retinopathy and associated risk factors in type 1 diabetes. Ophthalmology. 1998;105:1801-1815.

6. Gonzalez VUlalpando ME, Gonzalez Villalpando C, Arredondo Pérez B, Stern MP. Diabetic retinopathy in Mexico. Prevalence and clinical characteristics. Arch Mea Res. 1994:25:355-360.

7. Rabb MF, Cagliano DA, Sweeney HE. Diabetic retinopathy in blacks. Diabetes Care. 1990;13:12021206.

8. Kinyoun J, Barton F, Fisher M, Hubbard L, Aiello L, Ferris F III. Detection of diabetic macular edema. Ophthalmoscopy versus photography. Early Treatment Diabetic Retinopathy Study Report Number 5. The ETDRS Research Group. Ophthalmology. 1989-.96-.746-750.

9. Early Treatment Diabetic Retinopathy Study Research Group. Grading diabetic retinopathy from stereoscopic color fundus photographs: an extension of the modified Airlie House classification. ETDRS Report No. 10. Ophthalmology. 1991:98:786-806.

10. The Diabetes Control and Complications (DCCT) Research Group. The relationship of glycémie exposure (HbAIc) to the risk of development and progression of retinopathy in the diabetes control and complications trial. Diabetes. 1995:44:968-983.

11. Yang YS, Zou S, Zeimer R, Bressler NM, Schachat AP, Finkelstein D. Quantitative mapping of the retinal thickness in diabetic eyes with clinically significant macular edema. Invest Ophthalmol Vis Sd. 1997;38(siippI):S342).

12. Early Treatment Diabetic Retinopathy Study Research Group. Early photocoagulation for diabetic retinopathy. ETDRS report No. 9. Ophthalmology. 1991;98:766785.

13. Early Treatment Diabetic Retinopathy Study Research Group. Photocoagulation for diabetic macular edema. Early Treatment Diabetic Retinopathy Study Report No. I. Ana Ophthalmol. 1985;103:1796-1806.

14. Early Treatment Diabetic Retinopathy Study Research Group. Treatment techniques and clinical guidelines for photocoagulation of diabetic macular edema. Early Treatment Diabetic Retinopathy Study Repon No. 2. Ophthalmology. 1987:94:761-774,

15. Early Treatment Diabetic Retinopathy Study Research Group. Photocoagulation for diabetic macular edema: Early Treatment Diabetic Retinopathy Study Report No. 4, fnt Ophthatmol Clin. 193727:265-272.

16. Shahidi M, Ogura Y, Blair NP, Rusin MM, Zeimer R, Retinal thickness analysis for quantitative assessment of diabetic macular edema. Arch Ophthalmol. 1 99 1 ; 109:1115-1119.

17. Zeimer RC, Shahidi M, Mori MT, Benhamou E. In vivo evaluation of a noninvasive method to measure the retinal thickness in primates. Arch OphthalmoL 1989; 107:1006-1009.

18. Zeimer RC, Mori MT, Khoobehi B. Feasibility test of a new method to measure retinal thickness nonínvasively. Invest Ophthalmol Vis Sd. 1989;30:2099-2105.

19. Shahidi M, Zeimer RC, Mori MT. Noninvasive measurement of the retinal thickness. In: Wampler JE, ed. New Methods in Microscopy and Low Light Imaging. Bellingham, Washington: Society of Photo Optical Instrumentation Engineers; 1989:345-349.

20. Shahidi M, Zeimer RC, Mori M. Topography of the retinal thickness in normal subjects. Ophthalmobgy. 1990;97:1120-1124.

21. Zeger SL, Liang KY, Albert PS. Models for longitudinal data: a generalized estimating equation approach. Biometrics. 1988;44:1049-1060.

22. Vallortigara G, Bressan P. Occlusion, transparency, and stereopsis: a new explanation for stereo capture. Vision fos. 1994;34:289 1-2896.

23. Yang YS, Zou S, Zeimer R, Bressler NM, Schachat AP, Finkelstein D. Quantitative mapping of the retinal thickness in diabetic eyes with clinically significant macular edema. Invest Ophthalmol Vis Sd. 1997;38 (suppl): S342.

24. Zou S, Vitale S, Blankenship Jr, Yang YS, Zeimer R. Localized changes in retinal thickness after laser photocoagulation treatment for diabetic macular edema. Invest Ophthalmol Vis Sci. 1999; 40(suppl):S313).

25. Schaudig UH, Glaefke C, Scholz F, Richard G. Optical coherence tomography for retinal thickness measurement in diabetic patients without clinically significant macular edema. Ophthalmic Surg Lasers. 2000:31:182186.

Table 1. Characteristics of Patients*

Table 2. Association of RTI with Presence/Absence of Retinal Pathology Within 0.5 mm, Based on Areas Within Eyes (GEE Model to Account for Correlation of Areas Within the Same Eye)

Table 3. Association of RTI With Number of Retinal Pathologies Within 0.5 mm, Based on Areas Within Eyes (GEE Model to Account for Correlation of Areas Within the Same Eye)

Table 4. Association of Maximum RTI (of All Measured Areas Within an Eye) With Total Number of Retinal Pathologies in Posterior Pole, Based on Areas Within Eyes (GEE Model to Account for Correlation of Areas Within the Same Eye)

Table 5. Association of Foveal Mean RTI With Presence/Absence of Retina! Pathology, Eyes With at Least Some Retinopathy Anywhere in the Eye (GEE Model to Account for Correlation of Areas Within the Same Eye)

10.3928/1542-8877-20020301-09

Sign up to receive

Journal E-contents