M. Everett Lawson
RETINAL
IMAGING
Interaction as instrument. Perception as infrastructure.
Interaction Design
Perception as the interface
-
Gaze becomes input
-
Eye motion becomes navigation
-
The user completes the system through biological coupling
The interface isn’t on the screen. It’s in the body.
“Instead of asking the user to hold still, the system learns how to move with them.”
Perception Is the Interface
“The user doesn’t operate the system. The system listens to how the user sees.”
Designing With Biology, Not Around It
“The most powerful hardware in the system was already there.”
Cinematic UX
From a single perfect frame to a temporal narrative
-
No hero shot—only sequences
-
Meaning emerges over time
-
The retina is revealed cinematically through motion and montage
“A retinal image isn’t captured—it’s assembled.”
Motion Is Not Decoration — It’s Infrastructure
“When the eye moves, the experience improves.”
Interaction as perception, motion as meaning, and UX as a cinematic system unfolding over time.
It doesn’t ask what technology can display.
It asks how humans already see—and designs forward from there.
Motion Systems
Motion as structure, not decoration
-
Eye rotation drives spatial coverage
-
Gaze trajectories define capture paths
-
Temporal coherence replaces mechanical precision
Motion is the system.
“When the eye moves, the experience gets better—not worse.”
From Frames to Sequences
“There is no single image—only a story assembled over time.”
Challenge
Retinal imaging was built for stillness.
Chin on the rest. Eyes wide. Don’t move.
The retina—one of the most information-dense, light-absorbing tissues in the human body—has always demanded elaborate optics, trained operators, and perfect alignment through a pupil barely wide enough to admit a beam of light. The moment the eye moved, the system failed.
Self-imaging the retina wasn’t just difficult.
It violated the assumptions the field was built on.
Insight
Instead of immobilizing the eye, I realized we could collaborate with it.
Vision is already interactive. The eyes move together. Focus follows intent. Gaze is programmable through stimulus. If one eye is guided, the other complies—automatically, biologically.
The breakthrough wasn’t optical.
It was behavioral.
The retina didn’t need a single pristine image. It needed many imperfect ones—captured during natural eye motion—and a system smart enough to assemble them. Precision hardware could be replaced by perception, computation, and motion.
Execution
-
The system collapsed a room-sized medical instrument into something closer to eyewear.
-
A low-cost micro-camera was placed near the eye. Illumination was rethought entirely—using indirect, diffused LEDs through skin and programmable direct lighting tolerant to misalignment. A stimulus display guided gaze and focus via bi-ocular coupling while the camera captured fragments of the retina during eye rotation.
-
No moving parts.
No clinician.
No single “great shot.” -
Instead, computational photography selected, corrected, and mosaicked only the usable moments—turning eye motion into spatial coverage. A light-field variant pushed the system further, capturing retinal rays themselves and enabling refocus after capture.
-
For the first time, the retina became something you could interact with, not just observe.
Impact
-
The first self-directed retinal imaging system
-
The first near-eye system to use bi-ocular gaze coupling
-
The first successful light-field capture of the human retina
-
Beyond the technical milestones, it reframed retinal imaging as an experience design problem—one where optics, software, and human behavior are co-authors. It pointed toward a future of wearable diagnostics, accessible screening, and perceptually integrated sensing systems.
This invention comprises an apparatus for retinal self-imaging. Visual stimuli help the user self-align his eye with a camera. Bi-ocular coupling induces the test eye to rotate into different positions. As the test eye rotates, a video is captured of different areas of the retina. Computational photography methods process this video into a mosaiced image of a large area of the retina. An LED is pressed against the skin near the eye, to provide indirect, diffuse illumination of the retina. The camera has a wide field of view and can image part of the retina even when the eye is off-axis (when the eye’s pupillary axis and cameras optical axis are not aligned). Alternately, the retina is illuminated directly through the pupil, and different parts of a large lens are used to image different parts of the retina. Alternately, a plenoptic camera is used for retinal imaging.
In an initial alignment step, the imaging device displays real-time visual feedback to one eye (the stimulus eye) of a user. The visual feedback is indicative of (i) the pupillary axis of the user’s eye that is being imaged (the test eye) and (ii) the optical axis of the device’s camera. For example, an LCD in the device may display visual feedback that comprises a circle representative of the optic disc of the test eye (which serves as an approximate indication of the pupillary axis) and a square indicative of the center of the camera (which serves as an approximate indication of the optical axis of the camera). This real-time visual feedback guides the user as the user changes direction of gaze in order to self-align the two axes. Once the two axes are aligned, the imaging device displays a video of moving visual stimuli to the stimulus eye. The user’s stimulus eye tracks this moving stimuli.
Due to bi-ocular coupling, the test eye moves (rotates) in a similar path. As the test eye rotates into different positions, a camera in the device captures multiple images of different portions of the retina of the test eye. Each of these images may capture only a small portion of the retina. These images are processed and stitched together to form an image of a large area of the retina. This large field of view (FOV) image of the retina can be displayed to the user in real-time. As the test eye rotates (while bi-ocularly coupled to the stimulus eye), the test eye moves into many rotational positions in which the test eye is “off-axis’ with respect to the camera. As used herein: (i) an eye is “off-axis’ with respect to a camera if the optical axis of the camera is not pointed at the pupil of the eye; and (ii) an eye is “on-axis’ with respect to a camera if the optical axis of the camera is pointed at the pupil of the eye. The camera has a wide FOV and thus can capture an image of at least a small part of the retina, even when the test eye is off-axis.