Abstract
Abstract: :
Purpose:To investigate the feasibility of determining macular pigment optical density (MPOD) distributions from pairs of digital retinal images, and to compare the results with those of heterochromatic flicker photometry (HFP). Methods: A 3CCD, non–mydriatic retinal camera was modified by inserting 1) a triple–bandpass filter (460, 528, 610 nm) and 2) a double–bandpass filter (535, 670 nm) in the light path between the flash–lamp and subject's eye. Twenty–degree retinal images were obtained, without dilation or bleaching, from 8 subjects aged 18 – 25, 2 subjects aged 32, and one subject aged 60. Image analysis software was used to align the resulting images and extract 4 bitmaps representing the relative reflectance of the retina at 460, 528, 610 and 670 nm. (The 535 nm band was used to aid image alignment.) By linearly combining the corresponding log–transformed bitmaps, a bitmap representing the MPOD distribution was generated. (Three different linear combinations generated OD maps of the cone photopigments, rod photopigment and melanin.) MPOD for the subjects was also determined by HFP using a 1.5° stimulus. Results: As expected, MP, as well as cone photopigments and melanin distributions, peaked at the center of the macula, while the rod photopigment dipped in this region. Peak MPOD values for the ∼ 20 year old subjects were linearly correlated with values obtained by HFP (r2 = 0.76, p < 0.005). For the ∼ 20 year old subjects, HFP gave lower MPODs than those obtained by reflectometry, whereas for the 32 and 60 year old subjects, the opposite was found to be true. Conclusions: The method shows promise as a simple, fast, objective means of assessing a subject's MPOD. It eliminates the need to bleach the cone and rod photopigments as in traditional reflectometry. With older subjects, light scatter in the optic media may be responsible for an underestimate in MPOD by reflectometry. Eliminating this problem is the subject of current research efforts.
Keywords: macular pigment • carotenoids/carotenoid binding proteins • imaging/image analysis: clinical