10.2.6 Field Size
A large field size is almost always advantageous. A larger size permits easier
navigation through and across the retina and reduces certain complications
in postprocessing, such as those stemming from eye motion (e.g., image registration).
Several camera- and eye-related factors, however, constrain the
field size for AO imaging. These must be carefully weighed in context with
the application and are described below.
Retinal Safety An increase in field size, while maintaining constant retinal
illumination (J/cm2) as well as (CCD) pixel sampling density (pix/cm2),
requires that the total light flux entering the eye increase as the square of the
field size diameter (i.e., proportional to the retinal area being illuminated).
Under these conditions, the number of photons per pixel is held constant.
While the total light flux incident on the retina increases as the square of the
illumination diameter, the retinal hazard does not for intermediate field sizes
(1.5 to 100 mrad or 0.09° to 5.7°). At intermediate sizes, thermal heat generated
by the light dissipates less rapidly. Specifically, the thermal maximum
permissible exposure (MPE) grows linearly with (rather than the square of)
the field size from 1.5 mrad (0.09°) to 100 mrad (5.7°) [17]. This means that
the MPE in terms of retinal irradiance actually decreases in a linear fashion
with field size. Within this field range, the highest retinal irradiance therefore
occurs when using the MPE for the smallest field of 1.5 mrad. For diameters
larger than 100 mrad, the MPE grows as the square.
Isoplanatism The eye is an excellent wide-field imaging system but ocular
aberrations still change with the field angle. The isoplanatic angle refers to
the maximum angular extent over which the aberration pattern remains
largely constant. In a conventional system, AO corrects the center of the field
where it provides optimal correction. At field extents where the aberration
pattern has appreciably changed, the correction is effectively lost. High image
quality is restricted to within the isoplanatic patch created by the eye and
ophthalmoscope combination. The isoplanatic angle of the eye has not been
quantified, but in practice, images as large as 2° in extent have shown no
degradation in image quality at the edges of the field. Changes in the aberration
pattern beyond that field size may prevent effective AO correction, but
it remains to be explored.
Detector Size and Image Acquisition Rates Essentially all AO ophthalmoscopes
to date sample the retina at least two times finer than the diffraction-limited
PSF. This assures that all spatial frequencies up to the diffraction limit
cutoff are properly recorded (no aliasing). For large field sizes, this results in
a considerable pixel count. For example, given a diffraction-limited PSF with
a Rayleigh resolution limit of 1.9 μm [using Eq. (10.4) with λ0 = 0.55 μm, d =
6 mm, F = 22.2 mm, and n = 1.33] and a pixel size of 0.95 μm (at the retina),
approximately 2609 × 2609 pixels are required to sample an 8° field (~2400 μm
in diameter). If the images are 2 bytes per pixel and are sampled at video
rates, the data rate is ~410 MB/s, which is beyond the bandwidth of current
data storage systems.
Scattered Light Light can scatter at oblique angles as well as multiple times
in the thick retina, causing a fraction of the incident light to eventually exit
the retina at a location different from where it entered. This reduces the contrast
of retinal structures in the image. Smaller fields have been observed to
generate higher contrast images, suggesting that the amount of multiply scattered
light per retinal area that strikes the science camera (CCD) grows with
field size. For example, the highest contrast images of cones were obtained
with a small 7-arcmin illumination field [18].
10.2.6 Field Size
A large field size is almost always advantageous. A larger size permits easier
navigation through and across the retina and reduces certain complications
in postprocessing, such as those stemming from eye motion (e.g., image registration).
Several camera- and eye-related factors, however, constrain the
field size for AO imaging. These must be carefully weighed in context with
the application and are described below.
Retinal Safety An increase in field size, while maintaining constant retinal
illumination (J/cm2) as well as (CCD) pixel sampling density (pix/cm2),
requires that the total light flux entering the eye increase as the square of the
field size diameter (i.e., proportional to the retinal area being illuminated).
Under these conditions, the number of photons per pixel is held constant.
While the total light flux incident on the retina increases as the square of the
illumination diameter, the retinal hazard does not for intermediate field sizes
(1.5 to 100 mrad or 0.09° to 5.7°). At intermediate sizes, thermal heat generated
by the light dissipates less rapidly. Specifically, the thermal maximum
permissible exposure (MPE) grows linearly with (rather than the square of)
the field size from 1.5 mrad (0.09°) to 100 mrad...
More >>
