Retina or Cortex: How the Two Main Bionic Eye Strategies Differ, and Who Each One Is For

Bionic eyes are not one technology but two fundamentally different strategies, separated by where they plug into the nervous system. Retinal implants sit inside the eye and stimulate surviving retinal neurons after light-sensing cells have died; cortical implants skip the eye entirely and stimulate the brain's visual cortex. Which approach makes sense for a given person depends almost entirely on where their vision loss originated — a distinction that determines candidacy before any device, cost or surgical question even arises.

Understanding Bionic Eyes: Retinal vs. Cortical Approaches

The core organizing question for any bionic eye is simple: where is the damage, and what tissue is still healthy enough to carry a signal? A comprehensive review of the field describes retinal implants as devices designed to bypass damaged photoreceptor cells and directly stimulate the remaining retinal neurons, thereby creating artificial vision. That logic works only when the retina's downstream wiring and the optic nerve are still functional.

Cortical brain implants take the opposite path. Rather than rescuing a damaged eye, they aim to create artificial vision by directly stimulating the visual cortex, bypassing the eye and optic nerve altogether. That makes them, in principle, an alternative for people whose blindness originates beyond the retina — for example, optic nerve damage — though the evidence base for this approach remains far less developed than for retinal devices, and the sources reviewed here discuss cortical implants only in general terms rather than naming specific systems.

How Retinal Implants Restore Vision

Retinal implants exploit a biological fact: in diseases like retinitis pigmentosa, the photoreceptors die but other retinal cells survive. The implant substitutes for the lost photoreceptors by delivering electrical stimulation directly to those surviving neurons. In the epiretinal design used by the Argus II, the signal chain begins outside the body: an external camera captures images, the data is processed, and it is transmitted wirelessly to an implanted chip whose electrodes stimulate the retina's ganglion cells.

The Argus line has the longest clinical history. The Argus I Retinal Prosthesis, created by Second Sight Medical Products, was among the first retinal prostheses tested in clinical trials, entering human testing in 2002. Its successor, the Argus II, achieved basic shape recognition and motion detection, though at limited resolution. Other retinal technologies take different routes to the same target: the Alpha-IMS and Alpha-AMS devices and the PRIMA photovoltaic system represent subretinal approaches, while suprachoroidal devices sit between the choroid and the sclera — a placement researchers with Bionic Vision Australia explored with a 24-channel prototype.

Cortical Implants: A Different Path to Sight

For people whose retina or optic nerve is too severely damaged for a retinal implant, the cortical approach offers a theoretical way around the problem entirely. By stimulating the visual cortex — the part of the brain that normally receives signals from the eyes — these devices do not depend on any surviving ocular tissue. The eye and optic nerve are bypassed completely.

This independence is both the promise and the difficulty. Because the stimulation happens at the brain's end of the visual pathway, cortical implants could in principle serve patients no retinal device can reach. But the field is younger and the clinical evidence thinner than for retinal prostheses, and the sources available here treat cortical brain implants as a general research category rather than naming approved devices. Readers should therefore read this pathway as a direction of development, not an established clinical option comparable to retinal implants today.

Bionic Eye Technologies: A Comparative Overview

The table below sorts the major approaches by their point of neural interface — the single variable that determines candidacy. Read down the "Target Patient Condition" column to see why the same device cannot serve every cause of blindness: retinal implants require a healthy optic nerve and surviving retinal neurons, while the cortical route exists precisely for cases where that requirement fails.

Bionic eye technologies categorized by neural interface point
Technology Approach	Mechanism of Action	Example Device(s)	Target Patient Condition
Retinal implants (epiretinal)	External camera captures images; processed signals are transmitted to an implanted chip that directly stimulates remaining retinal neurons, bypassing damaged photoreceptors	Argus II (Second Sight)	Advanced retinal disease such as retinitis pigmentosa, where photoreceptors are damaged but the optic nerve and remaining retinal neurons are functional
Retinal implants (subretinal and photovoltaic)	Stimulation of surviving retinal cells from beneath or within the photoreceptor layer	Alpha-IMS, Alpha-AMS, PRIMA bionic vision system	Inherited retinal degeneration; PRIMA has also been studied in age-related macular degeneration contexts
Retinal implants (suprachoroidal)	Electrode array placed between the choroid and the sclera, stimulating the retina from outside the neural layers	Suprachoroidal–transretinal stimulation (STS) system; Bionic Vision Australia prototype	Retinal degeneration; a less invasive placement under investigation
Cortical brain implants	Direct stimulation of the brain's visual cortex, bypassing the eye and optic nerve entirely	Discussed in general terms; no specific devices named in the reviewed evidence	Blindness originating beyond the retina, such as optic nerve damage, where retinal implants are unsuitable

One caveat belongs alongside this comparison: the devices listed under retinal approaches are at very different stages. The Argus II is an early-generation system with documented long-term stability problems, while others — PRIMA, Phoenix 99, the IRIS II, and the STS system among them — remain in research or refinement, as Fighting Blindness Canada's overview of ongoing prostheses research makes clear.

Current Challenges and Future Directions

The honest summary of what today's bionic eyes deliver is: some vision, not restored sight. The comprehensive review characterizes Argus II outcomes as basic shape recognition and motion detection at limited resolution, and Firstpost's explainer similarly notes that current capabilities are basic, though steadily improving. For people going, as one clinician quoted by the American Academy of Ophthalmology put it, "from nothing to something, and that means a lot to them."

Three structural problems dominate the field's to-do list. First, longevity: the review documents electrode impedance increases and device degradation in the Argus II over time. Second, resolution: the number and size of electrodes constrain how much visual detail a device can convey, and as Eyes On Eyecare notes, the retina's limited anatomical real estate makes simply adding electrodes harder than it sounds — a constraint Science Corporation and Bionic Vision Technologies are both working on. Third, access: more large-scale clinical trials are required before these devices can be commercialized for the public.

The research direction follows directly from those limits. Ongoing work described by Fighting Blindness Canada aims to extend device lifespans, shrink and strengthen the hardware, expand the field of vision, increase electrode counts, and improve the clarity and sharpness of the perceived image. Firstpost adds a longer-horizon possibility: devices that manipulate images in ways the human eye cannot, or see beyond the visible spectrum.

Navigating Bionic Vision: What to Consider

When considering bionic eye technology, individuals should understand that the choice of implant strategy — retinal versus cortical — depends critically on the underlying cause and location of vision loss. The specific diagnosis of vision loss, such as photoreceptor degeneration versus optic nerve damage, determines the most suitable bionic eye approach. Someone with retinitis pigmentosa and an intact optic nerve falls squarely in retinal-implant territory; someone whose optic nerve is destroyed does not, and represents the population cortical research is meant to eventually serve.

Two boundaries matter for expectations. This article compares technological approaches; it is not medical advice and does not recommend specific devices. And clinical outcomes discussed here reflect published research, not guaranteed results for every patient. The measurable next milestone to watch is concrete: the large-scale clinical trials that, in the assessment of current reviewers, still stand between today's rudimentary-but-real artificial vision and broader commercial availability.