Framework
PAINT
The measurement framework: Position, Area, Intensity, Nature, Time.
How it works
Measure pain. See what it produces.
Position, Area, Intensity, Nature, and Time — captured on an anatomy-classified 3D surface, measured in real square millimetres, and exported as structured records.
The surface
A measured atlas, not a drawing of a body.
Painting happens on a 3D mannequin whose surface is divided into a fixed grid of texels and classified, triangle by triangle, against a reviewed anatomy atlas. Because the grid is fixed and the classification is reviewed rather than estimated, a painted region has a real area in square millimetres and a real anatomical label — and the same region on the same body variant means the same thing in every capture.
- 676,850 active paint texels across a 1024² atlas — 64.55% coverage.
- 28,092 individually classified anatomy patches.
- One engine paints, measures, and persists, so there is no gap between the live view and the saved report.
- Sampling resolution varies by body variant, because the same atlas wraps bodies of different sizes.
| Sampling | Male | Female | Youth / neutral |
|---|---|---|---|
| Resolution | 1.76mm | 1.64mm | 1.53mm |
| Mean texel area | 3.09mm² | 2.70mm² | 2.35mm² |
| Body surface area | 20,920cm² | 18,280cm² | 15,880cm² |
| Reference body | 1.97 m, 80 kg | 1.85 m, 65 kg | 1.65 m, 55 kg |
The framework
PAINT — five axes of a symptom record.
Four axes describe a symptom. The fifth is the one records almost always lose.
-
Position
Anchors each symptom to where it sits on the body — on a 3D model you can pose to match the patient, with that posture recorded alongside the chart.
-
Area
Measures how much body surface is involved, in real cm² and percent BSA on the reviewed 3D mannequin.
-
Intensity
Keeps symptom strength attached to the painted surface so the PAIN-Index is not flattened into presence versus absence.
-
Nature
Preserves the qualitative character of the symptom so burning, aching, pressure, or numbness remain distinguishable.
-
Time/ Trajectory
Tracks how a symptom changes across snapshots — whether it persists, spreads, resolves, or recurs — so reassessment asks where it is better, worse, new, or resolved.
The T in PAINT
Time is the letter. Trajectory is the idea.
Each visit is a PAINprint Snapshot. Stack snapshots and you get a PAINprint Timeline — the symptom story over time. Trajectory is how that story moves: spreading, shrinking, intensifying, resolving, or recurring. That is the difference between asking "where does it hurt?" and "where did it hurt, how much, what kind, and what changed?"
Single finding
PAINprint
The painted spatial record for one finding at one moment.
Timepoint / visit
PAINprint Snapshot
A time-stamped chart state composed of one or more PAINprints.
Longitudinal New
PAINprint Timeline
An ordered sequence of Snapshots — the symptom story over time.
Impact summary
PAIN-Index
A normalized impact score derived from measured area × reported intensity.
(weighted area / total BSA) × 100Derived objects
What each capture produces.
Beyond the exports, every capture yields a derived metric panel, a snapshot, and — stacked over time — a PAINprint Timeline. Examples are simulated and de-identified.
PAIN-Index panel
PAIN-Index ............ 7.4 / 100
Painted footprint ..... 3.1% BSA
Adjusted local intensity 6.8 / 10PAINprint Snapshot
Snapshot · Visit 3 · 2026-05-12
2 findings · right lumbar, right thigh
Replays to the same rendered surface.PAINprint Timeline
Timeline (3 snapshots)
new · resolved · spread · persistent
PAIN-Index 9.1 → 8.0 → 7.4Outputs
Six formats ship today.
Every output derives from one finding-level model. Examples are simulated and de-identified — they show the structure, not real patient data.
Encounter HTML
<section class="encounter">
<h1>Symptom map — visit 3</h1>
<p>Findings: 2 · PAIN-Index: 7.4</p>
<figure><img alt="3D body chart"/></figure>
</section>Timeline HTML
Baseline → Visit 2 → Visit 3
PAIN-Index 9.1 8.0 7.4
Area %BSA 4.2 3.6 3.1
Status: 1 finding no longer chartedPlain-text clinical note
Patient reports aching, burning pain over
the right lumbar region and posterior thigh.
Area ~3.1% BSA. PAIN-Index 7.4 (was 9.1).
Left arm no longer charted since baseline.Research CSV
finding,region,side,nature,area_cm2,pct_bsa,intensity
1,lumbar,right,aching,612,3.1,7
2,thigh,right,burning,344,1.7,6Canonical JSON
{
"schema": "painprint/3",
"snapshot": "2026-05-12T14:08Z",
"findings": [
{ "id": 1, "nature": "aching",
"areaCm2": 612, "intensity": 7 }
]
}FHIR R4 DocumentReference
{
"resourceType": "DocumentReference",
"status": "current",
"type": { "text": "Symptom map" },
"content": [{ "attachment": {
"contentType": "text/html" } }]
}Where the software stops
Every number here is arithmetic, not an opinion.
There is no model in this pipeline, no inference, and no prediction. Each figure is a deterministic transform of what a human put on the chart, and the exact contract that produced it — scoring version, body-model version, every formula, the area threshold, the body-surface-area denominator, the overlap policy — is emitted alongside the structured exports. That is what makes a number from two years ago still readable.
- Change between two captures is reported as a signed quantity and its direction.
- Whether that change matters is a clinical judgement, and the software does not offer one.
- Magnitude bands and detection thresholds are deliberately not shipped — they require minimal-detectable-change studies that have not been done.
- Given the published contract, any figure can be rederived by hand.
See it on a real chart.
The workspace is free and open, and nothing you draw leaves your device.