Industries / Health / Healthcare / Optometry
Eye exams
Eye exams use instruments and tests to look inside the eye and to measure how well it sees.
7
Milestones achieved, from 1851 to 2015.
1
To watch for: announced, not yet confirmed outside.
2
Open deliverables. Nobody has shipped them.
The road
AchievedWatch forOpen
Achieved
1851
Light into the living eye
Hermann von Helmholtz built an ophthalmoscope, an instrument that reflects a lamp's light into the eye off a glass plate and corrects the focus of the light coming back, so the examiner can see the retina and the head of the optic nerve. It let clinicians see them in living patients for the first time, a 2026 history in Frontiers in Medicine says. A 2024 review in Cureus dates his first version to 1850 and notes that Jan Purkinje had examined the back of the eye with a candle and short-sight spectacles in 1823, work his peers disregarded. The first International Congress of Ophthalmology, held in Brussels in 1857, recognized the instrument's importance, the same review says.
Credit
Hermann von Helmholtz
1862
Letters sized to an angle
Herman Snellen, an ophthalmologist in Utrecht, published letter charts in which each letter, at its stated distance, spans 5 minutes of arc, with strokes 1 minute wide. Acuity is written as a fraction, the test distance over the distance at which the smallest letter read spans 5 minutes, so 6/6 (20/20 in feet) is the usual standard. In 1864 the British military medical service ordered 1,000 copies, which helped spread them, Paulus de Jong wrote in Eye in 2022. In 2025 the UK College of Optometrists still called the chart synonymous with the eye test.
Lessoncopies drifted apart until, after A. G. Bennett complained in 1965 that charts were no longer comparable, I. L. Bailey and J. E. Lovie's 1976 design, with five letters of equal legibility on every row and sizes in steps of 0.1 log unit (the logMAR chart), became the basis of the ETDRS chart of 1982, now the standard in research.
pmc.ncbi.nlm.nih.govcollege-optometrists.orgeuropepmc.orgeuropepmc.org
Credit
Herman Snellen, working with Frans Donders
1911
A slit of light
Allvar Gullstrand, a Swedish ophthalmologist, invented the slit lamp, which throws a narrow, very bright rectangular beam into the front of the eye. Seen through a microscope, the beam lights a thin slice of the cornea and the lens. Gullstrand built it for his own research on how the lens focuses; clinicians took it over, and it is now used routinely to examine the eye, a 2013 history in the Irish Journal of Medical Science says.
Credit
Allvar Gullstrand
1950
Pressure measured by flattening
High pressure inside the eye is a main risk for glaucoma, a disease that damages the optic nerve. Hans Goldmann's applanation tonometer of 1950 measures the force needed to flatten a small, fixed area of the cornea; R. L. Stamper's 2011 history says it began the era of accurate pressure measurement and is still the most widely used tonometer in the world. Goldmann's instrument touches the cornea. Bernard Grolman of American Optical first showed a non-contact tonometer, which flattens the cornea with a puff of air and needs no anesthetic drops, at the American Academy of Optometry's 1971 annual meeting, according to the Optometry Museum of the Australian College of Optometry.
Lessonpressure alone does not settle a diagnosis; Stamper notes that a high reading, or a normal one, can no longer be counted on to diagnose glaucoma, though pressure still guides its treatment.
Credit
Hans Goldmann; Bernard Grolman and American Optical
1975
Field tests by computer
A visual field test maps where in the field of view a person can see a faint spot of light. Franz Fankhauser spent 15 years automating it: attempts with moving targets from 1958 failed, a switch to fixed spots brought a breakthrough in 1973, and the Octopus perimeter came on the market in 1975, B. P. Gloor wrote in Survey of Ophthalmology in 2009. The Humphrey Field Analyzer followed in the 1980s, and the two makes became nearly universal, the Imaging and Perimetry Society says. Faster testing methods, SITA from the late 1990s and SITA Faster in the 2010s, cut a standard test from about 15 minutes to about 2 minutes per eye, Sylvia Groth wrote in 2021.
Lessonautomation did not remove noise; the Imaging and Perimetry Society notes that results still vary widely from one test to the next, more so as vision is lost.
Credit
Franz Fankhauser (Octopus); Humphrey Instruments; Bengtsson and Anders Heijl (SITA)
1991
Cross-sections of the retina
Optical coherence tomography (OCT) builds a cross-section image from faint echoes of near-infrared light, much as ultrasound does with sound. David Huang, James Fujimoto and colleagues at MIT and Harvard reported it in Science in November 1991, imaging the retina of a donated eye at a resolution of a few micrometers. Humphrey (Zeiss) sold the first regulatory-cleared OCT in 1996. Only about 180 units had sold by 1999 and managers considered dropping it, but by 2006 more than 6,000 systems had sold and about 20 million scans had been done, by the estimate of Fujimoto and Eric Swanson.
Lessonan instrument that later became routine nearly died in its first years on sale.
Credit
David Huang, Eric Swanson, Carmen Puliafito, James Fujimoto and colleagues; Humphrey (Zeiss)
2015
Acuity on a phone
Peek Acuity is a smartphone app that shows a single tumbling E, a letter E turned to face one of four directions, and the person points the way it faces. In a 2013 to 2014 study of 300 adults aged 55 and older in Nakuru, Kenya, its designers found its scores within 0.07 logMAR of the ETDRS chart on average, repeatable to within ±0.033 logMAR, and no slower than a Snellen chart (77 seconds against 82); community health workers used it with little training.
Lessonagreement in adults did not carry over to children; an outside team from the University of Wisconsin and Fundación Visión found that in 393 children aged 6 to 16 in Paraguay the app referred only 48 percent of those who needed care, because it tended to overestimate their acuity.
Credit
Andrew Bastawrous, Matthew Burton and colleagues, London School of Hygiene and Tropical Medicine
Watch for
Field tests in a headset
PeriVision, a spin-off of the University of Bern, and its US partner VEO Ophthalmics say the VisionOne headset perimeter uses a learned test strategy (SORS) that cuts testing time by 70 percent while correlating strongly with the Dynamic Strategy, a standard test method. A 2026 study at Bern of 120 visual fields from 73 people with glaucoma found a correlation of 0.97 with an Octopus 900, but one author works for PeriVision and the others are at the university it came from.
Achieved whena group with no tie to PeriVision publishes agreement with a Humphrey or Octopus perimeter in its own patients.
Claimed by
PeriVision and VEO Ophthalmics, October 2025
Open
01
Acuity apps on many phones
A public measurement of how big the letters really are in free acuity apps and web acuity tests on at least 20 phones, tablets and laptops, Android included. Photograph each row on screen beside a ruler, compute the letter's height, and compare it with the height its stated line requires at the stated distance (5 minutes of arc for the whole letter). Record screen luminance too.
The briefCredit · unclaimed02
Field-test headsets on a bench
Measure, with a spectroradiometer placed where the eye would be, the light that commercial headset perimeters actually show for each decibel (dB) level they report, at the center and out to 30 degrees, and publish the gap between stated and measured values.
The briefCredit · unclaimed
Open 01 · the brief
Acuity apps on many phones
The first to publish is credited on this line, and the deliverable moves from Open to Achieved.
What to build
A public measurement of how big the letters really are in free acuity apps and web acuity tests on at least 20 phones, tablets and laptops, Android included. Photograph each row on screen beside a ruler, compute the letter's height, and compare it with the height its stated line requires at the stated distance (5 minutes of arc for the whole letter). Record screen luminance too.
Why it is still open
Two audits covered Apple devices only. Perera and colleagues (Eye, 2015) found optotype size errors of 4.4 to 39.9 percent across 11 iPhone Snellen apps. Steren and colleagues (JAMA Ophthalmology, 2021) measured apps on four Apple devices and found mean errors from 0.2 to 109.9 percent on iPhones and up to 398.1 percent on iPads; no app was both accurate and consistent across devices. A 2025 review of 44 free apps (Lentz and colleagues, JMIR mHealth and uHealth) rated quality on a rating scale and found few clinically validated, but did not measure the letters. A search of Europe PMC and the web (November 2026) found no measurement across Android handsets or browser tests.
Built with what exists
The devices; a camera on a tripod and a steel ruler; Python with OpenCV to measure letter heights in pixels; a luminance meter, against the College of Optometrists' recommended chart luminance of 80 to 320 cd/m² (College of Optometrists).
First step
Install the ten most-downloaded free acuity apps on one Android phone, photograph the 6/6 (20/20) row beside a ruler, and post the first table.
Done looks like
A public table of at least 30 apps or web tests on at least 20 devices giving each row's size error in percent and in logMAR, with the photos and code, and a second person's rerun on five devices that matches within 2 percent.
Open 02 · the brief
Field-test headsets on a bench
The first to publish is credited on this line, and the deliverable moves from Open to Achieved.
What to build
Measure, with a spectroradiometer placed where the eye would be, the light that commercial headset perimeters actually show for each decibel (dB) level they report, at the center and out to 30 degrees, and publish the gap between stated and measured values.
Why it is still open
Lee and colleagues (PLoS One, September 2025) measured three consumer headsets (HTC Vive, Oculus Quest 2, Pico Neo 3) in their own test scene. The lowest threshold they could reach was 16 dB on the HTC Vive and 20 dB on the other two, and brightness fell toward the edge of the view; a standard perimeter covers a range of four log units, or 40 dB. A 2025 review of 19 studies (Vera and colleagues, Vision) named limited dynamic range as a common weakness of simpler devices. Neither measured the commercial perimeters (Heru, VisuALL, VisionOne, Radius XR) running their own tests. The nearest is Groth and colleagues at Vanderbilt (Translational Vision Science & Technology, January 2023), who checked one VisuALL headset with a luminance meter at the center of the screen only, assuming the display was even across its width; its brightest spot measured near 105 cd/m², against a stated 120.
europepmc.orgpmc.ncbi.nlm.nih.govpmc.ncbi.nlm.nih.govwebeye.ophth.uiowa.edu
Built with what exists
A spectroradiometer (Lee found that handheld luminance meters read too high); a mount that holds its lens at the eye position; the headsets, run in test mode with no patient; Python for the tables.
First step
Borrow a current consumer headset and one commercial headset perimeter from a teaching clinic and repeat Lee's central measurement on both.
Done looks like
For at least five devices, a published table of measured luminance against stated dB at 0, 10, 20 and 30 degrees, with a method that a second lab repeats to within 1 dB.