Metamorphopsia
is a vision defect caused by a distorted retina in one or both
eyes. The inflicted patient perceives this condition
as a dynamic distortion the geometry of his environment. Metamorphopsia
transforms the patient's world into a perpetual Escher-like environment.
Mild horizontal distortions produce binocular depth perception errors
as evident in the attached Figure. More severe horizontal distortions
as well as vertical distortions cause ghosting or frank double vision.
The patient finds navigating stairs or non-uniform topologies a challenge.
The sudden onset of the distortion causes the greatest difficulties,
whereas if the distortion develops more slowly there may be some
adaptation to the sensory input such that the patient may not even
realize that a visual defect exists.
A tool to document metamorphopsia
has been developed as part of the multi-Axis
Vision Evaluation System (MAVES). With distortion data that is derived
from testing a patient with MAVES, the 3D environment can be reconstructed
from the patient's point of view. This process involves projection
of the 3D dataset onto each binocular 2D retinal plane. The planar
data is perturbed by the documented distortions and two 3D datasets
are reconstructed. Visualization of the reconstructed datasets allows
a third party, such as the ophthalmologist or family member, to observe
the perturbations the patient sees.
The MAVES testing apparatus consists of a 21" video monitor
and PC, a chin-rest to fix the patient's head, an interaction control
pad, and fixation monitoring via an infra-red pupil tracker. The
metamorphopsia analysis program presents 5x5 regular spline grid
with random perturbations overlaid on a cross, which the patient
fixates at its center. The display grid dims if the pupil tracker
detects a fixation deviation. The patient is instructed to remove
the distortions from the grid until all of the lines appear straight
and uniform. He accomplishes this by manipulating over the grid
using buttons on the control pad and adjusting the deviation of
the lines using a joystick. The resulting grid objectively documents
the macular distortions.
The distortion data from the resultant grid may be applied to
a flat, 2D image. The distortion maps the image from the world
image
space to the patient's image space. An image with the distortion
applied will appear geometrically correct when presented to the
patient. The inverse distortion maps the image from the patient's
image space to the world image space. An image with the inverse
distortion applied is representative of the patient's monocular
vision aberration and has been extremely useful in demonstrating
to the ophthalmologist or family, the monocular distortion perceived
by the patient. However, the 2D image fails to adequately express
the real world 3D distortion the patient perceives with his binocular
vision.
A CG 3D environment may be reconstructed to demonstrate what
is perceived by the patient under binocular conditions. This process
involves projecting the 3D point coordinates of the data set onto
each binocular 2D retinal plane through each eye's nodal point.
The 2D planer points are perturbed by the documented distortion
measured for each eye. Two vectors defined by the perturbed 2D
points in each retinal plane and corresponding nodal point are
projected back into the world space occupied by the original geometry
to create two new 3D geometry datasets. The position where
the two vectors are closest is analyzed. If the distance between
the vectors (d) is smaller than a convergence error limit (k) then
the center point of the line drawn between the two vectors defines
the new 3D point for both geometry data sets. If d>k then two
new 3D points are defined by the point of closest approach for
each vector.
The two 3D geometry sets are rendered to produce 2 images, one
for each eye. A composite of both images is generated with 50%
transparency to produce a single image representative of the
patient's environment. As the patient moves, new geometry sets
are calculated
and rendered.
The warped 3D geometry appears to realistically represent the
visual disturbances which these patients describe and for the
first time
has given clinicians an insight into this poorly understood
and poorly appreciated aberration of vision.
www.clearvison.org
MAVES Siggraph Slides |
]
[
