Vision Science & Ophthalmology
A transmissive corrector close to the eye brings cellular-level resolution to existing ophthalmoscope designs.
Cellular-level retinal imaging
The eye aberrates its own image.
Cornea, lens and tear film distort the wavefront on the way in and again on the way out, and that distortion is why a clinical fundus image stops short of the individual photoreceptors beneath it. Adaptive optics takes it out: AO ophthalmoscopes resolve cones, and have largely stayed in research labs.
What keeps them there is the corrector. A deformable mirror has to be folded into the beam and conjugated to the pupil through relay optics, and that chain sets the instrument's footprint, its alignment tolerance and its cost. The Deformable Phase Plate is transmissive. It sits in line, close to the eye, and corrects the ocular wavefront in the path where it arises — no fold, no relay, no bench built around the mirror. For an instrument builder that is the difference between a corrector that can be added to an existing design and one that dictates a new one.
Case study
Dr. Pedro Mecê, Dr. Kate Grieve and Dr. Maxime Bertrand run a multimodal full-field OCT ophthalmoscope built for exceptional imaging speed and field of view. Its optical quality, though, was capped by the aberrations of whichever eye sat in front of it — and a conventional deformable-mirror upgrade would have cost more, and taken more space, than the instrument had to give.
They took the refractive route instead, and it came down to a single hardware change: a DPP after the final lens, in the imaging aperture facing the patient's eye. No relay optics, no folded beam path, no wavefront sensor. The existing housing already took standard cage components, so the mechanical work was a matter of minutes rather than an optical redesign, and full hardware and software integration was finished inside a day.
From there the instrument measures itself. Aberrations are estimated by modal decomposition from its own spectral-domain OCT data, so neither a Shack–Hartmann sensor nor a guide star is needed — one estimation step, about three seconds, before acquisition.
Apart from those three seconds the clinical workflow was untouched. Photoreceptors resolved individually, signal-to-noise doubled, and Yellott's ring — the signature frequency of the photoreceptor mosaic — appears in the Fourier transform only once the correction is on.
Let’s start with your instrument
Every patient is a different objective lens
In ophthalmoscopy the eye itself is the front optic. The pupil sets the numerical aperture; the cornea and the crystalline lens set the wavefront. Both differ from patient to patient, and in the same patient from one session to the next. There is no fixed specification to design around, so the correction has to be measured and applied for each eye as it sits in front of the instrument.
The corrector can sit at the front, facing the eye
A fully dilated pupil is about 8 mm, so a 10 mm clear aperture covers it. The corrector can therefore go straight behind the final lens, where the beam is already the size of the pupil — no relay, no conjugate pupil plane built deeper into the instrument. In practice the question is simply what else already occupies that space in your design.
How much retina do you need at cellular resolution?
One correction holds only over one isoplanatic patch, which is why adaptive-optics retinal fields are a few degrees across. To cover more, the instrument corrects one region, images it, then re-corrects for the next and stitches the results — more acquisitions in exchange for a clinically useful area at cellular resolution, rather than a single patch of it.
What is left after sphere and cylinder?
A trial lens or the instrument’s own focus takes out defocus and astigmatism — the patient’s prescription. What stays behind is coma, trefoil and spherical aberration, and it is those higher-order terms that keep the photoreceptor mosaic blurred. They are what the DPP is for, and it corrects them alongside the prescription rather than instead of it.
If any of them describes your instrument: how much of the retina are you still not resolving?
Your application is not on this list?
Let us build a customized DPP product based on your requirements to realize your vision of a novel way to apply adaptive optics.