> Background and Methods > Corneal Curvature Measurement with IR Keratometry: Principles and Workflow

Methods and Protocols

Optical methods and eye growth

Corneal Curvature Measurement with IR Keratometry: Principles and Workflow

Infrared keratometry uses a ring of eight IR LEDs reflected from the anterior corneal surface to measure the radius of corneal curvature as an indirect proxy for eye growth and structural change. This article explains the optical principle behind the Striatech Keratometer, details a practical measurement workflow for alert animals, and discusses how keratometric data complement refractive state measurements in myopia research.

Table of Content

Placeholder element - will be replaced automatically with TOC (h2 elements in the text)

Concepts and Approaches

Corneal Curvature as a Structural Marker of Eye Growth

Myopia development in animal models involves coordinated changes in multiple ocular structures: the axial length increases, the vitreous chamber deepens, and the cornea often flattens as the eye grows to a larger diameter. Measuring the radius of corneal curvature provides a non-invasive structural readout that can be repeated in alert animals at regular intervals, enabling researchers to track anatomical progression alongside functional optical changes. The Striatech Keratometer, developed by Prof. Frank Schaeffel at the University of Tübingen, captures the anterior corneal reflection of a ring of 8 IR LEDs and converts the ring image geometry to a curvature radius in millimeters.

Unlike photorefraction, which measures the integrated refractive error of the whole optical system, keratometry isolates the corneal surface specifically. When used together, the two measurements allow the investigator to ask whether a given diopter shift originates primarily at the cornea or deeper within the eye (Jiang et al., Advanced Science (2024)).

Why Corneal Curvature Changes with Eye Growth

In a spherical eye model, the anterior corneal radius of curvature is geometrically coupled to overall eye size: as the eye elongates and its equatorial diameter increases, the cornea tends to flatten (larger radius of curvature). In form-deprivation and lens-induced myopia models across chickens, guinea pigs, mice, and rats, measurable corneal flattening accompanies axial elongation, though the relative magnitude of the corneal change versus the axial change varies by species. In chickens, substantial corneal flattening is observed alongside rapid axial growth. In guinea pigs, corneal changes are moderate and contribute to the overall refractive shift (Cai et al., Experimental eye research (2026)). In mice and rats, axial changes are smaller in absolute terms, and the corneal contribution relative to lenticular and vitreous changes is an active area of investigation.

How the IR LED Ring Reflection Encodes Curvature

The Striatech Keratometer positions eight IR LEDs at equal angular intervals around the camera objective. These LEDs project a ring of light points onto the anterior corneal surface, which acts as a convex mirror. The reflected ring image is captured by the camera. The diameter of the reflected ring on the sensor is directly related to the corneal radius of curvature: a flatter cornea produces a larger reflected ring, while a more curved cornea produces a smaller reflected ring. The software fits a circle to the reflected LED positions and derives the curvature radius from the fitted circle geometry and the known LED-to-cornea distance. The measurement takes place on the anterior corneal surface only; posterior corneal curvature and internal surfaces are not assessed by this technique.

Step-by-Step Keratometry Protocol

The following workflow describes a standard longitudinal measurement session. As with photorefraction, parameters below are illustrative and should be adapted to species and experimental design.

Preparation. Power the Keratometer.
Animal preparation. Bring the animal to the measurement area and allow 1-2 minutes of acclimatization in a quiet environment. No anesthesia, mydriasis, or restraint is required for routine keratometry in alert animals. If the animal was previously fitted with a goggle or lens, remove the apparatus, allow any goggle-induced corneal indentation to resolve for a few minutes if present, then proceed.
Positioning and alignment. Hold or allow the animal to sit naturally on the platform. Position the instrument so that the camera axis is aligned with the corneal apex of the eye to be measured. The real-time display shows the LED ring reflection; a circular, symmetric ring centered in the pupil indicates correct axial alignment. Slight lateral misalignment causes the ring to shift toward one edge; correct before capturing.
Baseline session (day 0). The software records the corneal curvature with a high measurement interval. The largest source of measurement error comes from varying the distance between the camera and the cornea. Make sure that the reflected LEDs appear crisp and in focus. Record the mean curvature radius for each eye and note the standard deviation as a quality indicator. Document the time of day.
Longitudinal follow-up sessions. Perform keratometry at the same intervals as photorefraction sessions so that paired optical and structural data are available at each time point. Use the same operator and working distance at each session. The interocular difference in corneal radius is the primary structural outcome in unilateral induction studies, paralleling the refractive interocular difference measured by the Photorefractor.
Post-induction measurements. Continue measurements after the induction period ends to characterize any structural recovery. In species with rapid emmetropization (chicks, guinea pigs), corneal changes may partially reverse during recovery, providing a structural correlate of the functional refractive recovery.
Data export and quality control. Export session files for each time point. Flag sessions that fail quality criteria as missing data. For each animal, inspect the longitudinal trace of curvature radius to identify implausible jumps that may indicate operator error or inadvertent mechanical contact with the cornea.

For the complementary refractive endpoint, see the photorefraction protocol for alert animals.

Interpreting Keratometric Data and Detecting Artifacts

A progressive increase in corneal radius (flattening) in the treated eye relative to the fellow eye is the expected structural signature of form-deprivation or lens-induced myopia in most species. Stable or symmetric corneal radii in both eyes suggest the induction paradigm has not produced the expected structural response, warranting verification of apparatus placement and animal compliance.

Common measurement artifacts include: (1) incomplete LED ring due to eyelid encroachment at the upper or lower limbus, which reduces the number of fitted LED points and degrades curvature accuracy; (2) tear film irregularities immediately after blinking, which distort the reflected ring momentarily; (3) excessive head tilt, causing the ring to appear elliptical even on a spherical cornea. The within-session standard deviation across frames is an effective first-pass quality indicator. A standard deviation considerably higher than the typical instrument noise floor warrants rejection of the session or re-measurement.

Combining Keratometry with Photorefraction and Other Endpoints

Keratometry and photorefraction are designed as complementary tools (Jiang et al., Nano Energy (2024)). Recording both at each time point allows partitioning of the measured refractive change into a corneal component (estimated from the change in corneal power derived from the curvature radius, using the paraxial keratometer formula with appropriate refractive indices) and a residual non-corneal component. The residual primarily reflects axial elongation and lenticular changes. This partitioning is informative when testing compounds that may act on specific tissues: for example, a scleral remodeling agent might be expected to primarily affect the axial component, while a compound acting on the corneal epithelium might have a predominantly keratometric signature.

For a side-by-side comparison of what each instrument measures and guidance on when to use one or both, see Photorefractor vs Keratometer: complementary tools for myopia research.

Limitations and Boundary Conditions

Keratometry measures only the anterior corneal surface curvature. Changes in posterior corneal curvature, crystalline lens power, or vitreous chamber depth are not captured. In small-eyed species such as mice, the absolute difference in corneal radius between emmetropic and myopic eyes may be small relative to measurement noise, requiring larger sample sizes or more time points to achieve statistical power. In chickens and guinea pigs, the larger eye size and more pronounced corneal changes make keratometry more sensitive.

The keratometer assumes a spherical or mildly astigmatic anterior cornea for its primary curvature calculation. Eyes with substantial corneal irregularity, surface pathology, or prior surgical intervention may not satisfy this assumption, producing unreliable radius estimates. In such cases, the four semi-meridian values should be inspected individually before accepting the mean as representative.

Key Takeaways

The Striatech Keratometer measures the radius of corneal curvature in millimeters by analyzing the reflected image of 8 IR LEDs from the anterior corneal surface in alert, unsedated animals.

Compatible species include mice, rats, guinea pigs, and chickens.

Corneal flattening (increasing radius) in a treated eye relative to the fellow eye is the expected structural marker of experimental myopia in most species and induction paradigms.

Keratometry measures the anterior corneal surface only and cannot detect axial elongation or lenticular changes directly.

Pairing keratometry with photorefraction at each time point enables partitioning of overall refractive change into corneal and non-corneal components.

Sessions with high within-session standard deviation or incomplete LED ring capture should be flagged and either re-measured or excluded before analysis.

In mice, absolute corneal radius differences between myopic and emmetropic eyes are small, requiring adequate sample sizes and sensitive analysis plans.

Answers to the most frequent questions

What does the Striatech Keratometer measure and how does it differ from measuring refractive state?

Answer

The Keratometer measures the radius of curvature of the anterior corneal surface in millimeters by analyzing the reflected image of 8 IR LEDs. Refractive state, measured by the Photorefractor, captures the integrated optical power of the whole eye in diopters. Corneal curvature is a structural parameter reflecting the shape of one optical surface, while refractive state is a functional parameter combining contributions from all surfaces and the axial length. For guidance on using the two together, see Photorefractor vs Keratometer: complementary tools for myopia research.

How many IR LEDs does the Striatech Keratometer use and why is that number significant?

Answer

The Striatech Keratometer uses 8 IR LEDs arranged in a ring around the camera objective. Eight equally spaced reflection points provide sufficient angular sampling to fit a circle accurately to deduce spherical curvature.

Does keratometry require anesthesia or pupil dilation in the measured animal?

Answer

No. The Striatech Keratometer is designed for alert, unsedated animals without pupil dilation. The LED reflections originate from the corneal surface, not from the retina, so pupil size does not directly limit the measurement. Anesthesia helps to keep the animal stable; on the othe rhand, it could suppress natural head posture and or alter corneal hydration, potentially affecting the reflection geometry.

Which species show the largest and most detectable corneal curvature changes during experimental myopia?

Answer

Chickens and guinea pigs typically show the most pronounced and readily detectable corneal flattening during form-deprivation and lens-induced myopia induction. Mice and rats have smaller eyes and smaller absolute curvature changes, which makes their detection more dependent on sample size and measurement precision. Species selection for keratometry-dependent studies should account for this difference in effect magnitude.

What is the interocular difference in corneal radius and why is it the preferred outcome metric?

Answer

In unilateral induction paradigms, the interocular difference is the curvature radius of the treated eye minus that of the fellow (control) eye at a given time point. Using the fellow eye as a within-animal control subtracts out individual variation in baseline corneal size and any symmetric developmental changes, making the interocular difference more sensitive and specific to the induction effect than absolute values in either eye alone.

Can keratometry detect whether a goggle or lens is causing unintended mechanical pressure on the cornea?

Answer

To some extent, yes. Mechanical indentation from ill-fitting apparatus can produce localized corneal flattening that may appear as an unusually large change in corneal curvature. If such artefacts are detected, the fit of the induction apparatus should be re-evaluated before continuing the study.

How are keratometry and photorefraction measurements coordinated within a longitudinal study session?

Answer

The two measurements are typically performed at the same session for each time point so that paired structural and functional data are available. There is no fixed order requirement, but removing any induction apparatus before either measurement and restoring it immediately after ensures that the optics of the eye are in their natural state during both readings. Session timing, operator identity, and working distances should be standardized across all time points to minimize systematic drift.

What should I do if the standard deviation across frames in a keratometry session is unexpectedly high?

Answer

High within-session standard deviation most often reflects animal movement, incomplete LED ring capture due to partial eyelid closure, or tear film disruption after blinking. Allow the animal a brief rest period, ensure the room is quiet, and repeat the measurement. Sessions that cannot be resolved to an acceptable variance should be flagged as missing data rather than averaged with the high-variance frames.