[Detection of macular foveal by optical coherence tomography and its influence on results of measurement of central retina thickness].

ABSTRACT

OBJECTIVE: To detect macular foveal by Cirrus(TM) HD-OCT using the post acquisition fovea detection and to study its influence on the results of measurement of central retina thickness (CRT). METHODS: In this retrospectively studies, the 512×128 cube scans from 50 eyes of normal subjects and 50 eyes of macular edema patients were analyzed. Three methods for the foveal detection were used, the scan center (SC), the automated detected foveal (AF) and the manually identified foveal (MF). The foveal detected by MF was set as the reference. Radial distances from SC and AF to MF were calculated. Subfields retinal thickness acquired from different groups using different foveal detection methods were compared by paired Student's t test for normal data or Wilcoxon-Mann-Whitney for non-normal data. RESULTS: For normal subjects, it is reliable determining the foveal by special tomography features. The means of the radial distance from the MF was (35.0 ± 23.0) µm in scan center group and was (6.0 ± 9.0) µm in automated center (t = -3.732, P < 0.05). The average CRT difference between the SC and MF groups was (7.4 ± 10.4) µm. The median difference between the AF and MF groups was 0 µm (0 to 3), which is significantly less than that between the SC and MF groups (Z = -2.956, P < 0.05). In subjects with macular edema, the means of the radial offset was (59.0 ± 26.0) µm in SC group and was (34.0 ± 23.0) µm in AF group (t = -4.555, P < 0.05). The average CRT difference between the SC and MF groups was (28.0 ± 36.0) µm, which was not significantly greater than the difference between the AF and MF groups (23.0 ± 31.0) µm in macular edema eyes. CONCLUSIONS: Post acquisition detection of foveal reduces the radial offset to the real fovea, and may improve the reproducibility and reliability of the central retinal thickness quantification using spectral domain OCT in normal subjects.