WoundScribe AI
Share
X Facebook WhatsApp Email

Inside the healing trajectory agent for wound trial run-in periods

clinicaltrials

Published

How the healing trajectory agent screens DFU trial candidates during run-in, flags 20% shrinkage, and surfaces stalled wounds for enrollment decisions.

Run-in periods make or break wound trial enrollment. Enroll a fast healer and you dilute your effect; miss a truly stalled wound and you turn away a valid candidate. The healing trajectory agent is built for exactly this decision.

What the agent does during run-in

During the 2- to 4-week screening window, every visit produces a measured image. The healing dashboard computes percent area reduction (PAR) visit-over-visit and cumulatively from baseline.

  • Baseline area locked at screening visit 1
  • PAR calculated at each subsequent visit
  • Threshold alert (default 20%, configurable per protocol) when the wound has healed enough to fail run-in
  • Stalled flag when PAR is below expected trajectory

Why the timing matters

A wound that hits 22% PAR at the end of run-in should be excluded. Catching that on the day of the visit — instead of during a monitor query weeks later — protects both the participant's time and the trial's integrity.

Tied to signed source, not spreadsheets

Every data point feeding the trajectory comes from the AI-powered wound imaging agent and the investigator-signed note. There's no separate research-only measurement to reconcile.

Beyond run-in: mid-trial stall detection

Once enrolled, the same agent watches for stalls that trigger protocol-defined actions — additional imaging, adjunctive therapy, or discontinuation. Sponsors get an interpretable timeline; investigators get a heads-up before week 4.

Configurable per protocol

Bring your sponsor's PAR threshold, run-in length, and stall definition. The dashboard adapts. Start free and we'll map it to your protocol.