Competitive Brief · Clinical Pupillometry
Auditable by design
Objective pupillometry software for clinical and research neurology — measured against the dedicated hardware instruments and mobile apps clinicians compare it to.
| PupilMetrics NeuroSoftware · CNRI | NPi-300NeurOptics | PLR-3000NeurOptics | ReflexBrightlamp | RAPDx / EyeKinetixKonan | |
|---|---|---|---|---|---|
| Form factor | Software — desktop + phone/tablet, any UVC/webcam | Dedicated IR handheld | Dedicated IR handheld | iPhone app | Benchtop pupillograph |
| Primary market | Clinical + research neurology | Critical care / neuro-ICU | Research / clinical | Concussion screening | Ophthalmology / RAPD |
| Headline metric | Full PLR panel + PIPR, age-normed | NPi (0–4.9 index) | NPi + T75 | Latency, constriction rate | RAPD magnitude (log units) |
| Single-number index | PMi — published formula | NPi — proprietary | NPi — proprietary | No | No |
| T75 recovery | Yes | Extrapolate only | Yes | No | No |
| PIPR / melanopsin | Yes | No | No | No | No |
| Anisocoria | Yes | Manual, 2 scans | Manual | Limited | Yes (binocular) |
| RAPD | No (monocular) | No | No | No | Yes (core) |
| Drug-effect monitoring | Dedicated module | No | No | No | No |
| Serial trend tracking | Built-in | Via SmartGuard / EHR | Limited | App history | Yes |
| Regulatory | Research-use / licensed software | FDA cleared, CE | FDA cleared | Class I 510(k)-exempt | FDA cleared |
| Approx. cost | Software license — no proprietary hardware | ~$4–5k+/unit + disposables | ~$4–5k+/unit | $39.99/mo · $399.99/yr | Capital equipment |
PMi = 5·e−0.35 z̄
A single 0–5 score in the shape clinicians already read at the bedside — but every term, weight and age-norm is published in plr_pmi.dart. Each PLR parameter becomes a directional z-score against age-adjusted references, penalized only toward the pathological side, then combined as a weighted mean and mapped to the scale. Not equivalent to any cleared device’s index — the shape is familiar on purpose, the math is open on purpose.
Missing terms drop out and the surviving weights re-normalize; a poor-quality capture withholds the score entirely. Every contribution is returned for term-by-term inspection. PMi-1.0.0-experimental — unvalidated, not a substitute for any cleared index; research / illustrative use only.