CNRI PupilMetrics

Clinical Neuro-Optic Research Initiative

The pupil is a two-millimetre window onto the brainstem.

CNRI builds instruments that read it — the pupillary light reflex frame by frame, and the static shape of the pupil margin — and write down what they measured, in units, with the assumptions attached.

Everything runs on the device in front of the patient. No cloud, no account, no telemetry — and no patient record that leaves the room.

A pupillary light reflex trace Pupil diameter in millimetres over ten seconds. The pupil holds a six-millimetre dark-adapted baseline for three seconds, a two-hundred-millisecond flash fires, and after a two-hundred-and-fifty-millisecond latency the pupil constricts thirty percent to four point two millimetres before redilating, recovering seventy-five percent of the amplitude in one point four six seconds. 4 5 6 0s 2s 4s 6s 8s 10s 3 s dark baseline redilatation & PIPR 250 ms latency 30.0% constriction T75 = 1.46 s 6.00 mm dark-adapted mm
  • Pupil diameter sampled every frame, five-point median smoothed
  • Dark-adapted baseline the reference every amplitude is measured from
A normal trace. Every number the callouts print is measured back off this curve by the generator that drew it — latency 250 ms (normal 200–300), constriction 30.0% (normal 20–40), peak velocity 3.37 mm/s (normal 3–8). The curve is a model, not a recording; see what we have and have not validated.
Applications
Research · Neuro · Drug Monitor
Platforms
Windows · Android · macOS · iOS
Languages
18 languages, 20 locales
Patient data
Local to the device. Always.
02 — Why the pupil

The reflex arc runs straight through the midbrain

Light reaches the retina, the signal crosses to the pretectal nucleus, and two efferent limbs — one parasympathetic, one sympathetic — set the size of the pupil against each other. Every one of those relays sits somewhere a clinician would like to be able to see.

The pupillary light reflex arc A schematic loop: light enters the eye and travels the afferent limb through the optic nerve to the pretectal nucleus, then returns along the parasympathetic efferent limb through the Edinger-Westphal nucleus and the third cranial nerve to the pupillary sphincter, constricting the pupil. afferent · CN II pretectal nucleus Edinger– Westphal efferent · CN III → sphincter
The pupil in this diagram is not decorative — it runs the same constriction and recovery the trace above it describes.

What disturbs the arc shows up as a number

Direct brainstem injury. The pretectal and Edinger–Westphal nuclei lie in the midbrain, exposed to the rotational acceleration that is the hallmark of concussion. Even mild diffuse axonal injury slows PLR latency before clinical signs appear.

Rising intracranial pressure. Herniation compresses CN III. The earliest sign is a sluggish, asymmetric reflex — measurable before the pupil looks abnormal to the eye.

Autonomic dysregulation. Disruption of the hypothalamic–brainstem network produces abnormal resting size, delayed sympathetic redilatation and increased pupil unrest.

Loss of cortical modulation. The frontal lobe normally damps the reflex across repeated stimuli. Losing that input produces either excessive habituation or paradoxical sensitisation — which is what the three-trial protocol is for.

03 — Three trials, not one

One flash is a measurement. Three are a pattern.

In a healthy, unmedicated brain the reflex damps slightly across repeated identical stimuli. The size of that damping is the habituation index, and CNS-active drugs push it in class-specific directions — which a single trial cannot see at all.

Three PLR trials overlaid Three pupil traces on one timeline. Constriction amplitude falls slightly from the first trial to the third, giving a habituation index of twelve percent, inside the normal band. 4 5 6 0s 2s 4s 6s 8s 10s HI = (T1 − T3) / T1 = 12.0% 0–15% · normal cortical modulation
  • Trial 1 full amplitude, the reference
  • Trial 2 after two minutes' rest
  • Trial 3 the amplitude the index is computed from
Trials are separated by a standardised two-minute rest — the minimum for full sympathetic recovery back to a dark-adapted baseline.
Habituation index
HI rangeInterpretation
NegativeSensitisation — a paradoxical increase across trials
0–15%Normal cortical modulation
15–30%Moderate habituation — warrants monitoring
>30%Excessive habituation — disrupted cortico-midbrain feedback

Thresholds are taken from the apps' own built-in manual, so the site and the software say the same thing. The Drug Effect Monitor is this protocol with a longer flash and an extended recovery window.

04 — The initiative

Who we are and what we are trying to settle

Mission

CNRI advances the science of pupil-based neurodiagnostics by preserving historic clinical insights, developing modern analytic tools, and conducting rigorous research into the relationship between ocular micro-structures and systemic health. Our aim is to validate and expand neuro-optic biomarkers that can support early detection, monitoring and non-invasive assessment of autonomic and neurological function.

Vision

A future in which standardised neuro-optic biomarkers complement mainstream clinical assessment — earlier detection of imbalance, more personalised care, and a deeper understanding of the body's autonomic responses.

We want to be the institution that bridges historical clinical investigation and modern scientific rigour, so that decades of prior work are not lost but elevated, validated and expanded with contemporary technology.

Where the work comes from

The observational base is a body of clinical work carried out in Russia and Korea through the 1980s and 1990s. CNRI brings modern imaging, pattern recognition and data-driven method to it, to investigate how far subtle pupil irregularities do or do not reflect autonomic and neurological function.

We conduct non-invasive human studies, develop the analytic frameworks, and collaborate with researchers internationally.

What that inheritance does not buy

A historical association is a hypothesis with a long paper trail, not a validated finding. The zone-to-organ correspondences in PupilMetrics Research are labelled in the software as research-derived theoretical frameworks, and they are reported that way here too.

None of our instruments is a cleared medical device. Nothing in them has been validated against a reference pupillometer. The list of studies that would be needed is published rather than glossed.

Do you have a human pupil research project?

We collaborate on study design, provide instrumented builds with machine-readable export, and are interested in anyone in a position to run the method-comparison work our own instruments still need.