PIXAIRE-Redux study: validation of the Propose–Correct–Validate workflow for AI-based wound measurement with WoundTrack 2

A prospective validation study of the PCV (Propose–Correct–Validate) workflow applied to the WoundTrack 2 module of Pixacare, designed to measure both wound surface area and tissue composition (granulation, fibrin, necrosis) of chronic wounds, with a clinician validation step at its core.

Study framework

Study conducted using data from the Pixaire-1 cohort, with a comparative analysis based on the new methodological approach of the WoundTrack 2 module. Manuscript submitted for publication, currently under peer review.

Timeline

• Cohort reused: Pixaire-1 (enrollments from May to June 2023)
• Status: study completed, manuscript submitted and under peer review
• Expected publication: 2026

Study design

Analytical validation study comparing the WoundTrack 2 module to the reference method (digitized planimetry) on 38 chronic wounds, each assessed independently by two experts.

Methodological innovation: the PCV workflow

The study validates a new approach developed from the lessons of Pixaire-1, the PCV workflow (Propose–Correct–Validate):

• Propose: the algorithm proposes an automated segmentation of the wound from the photograph
• Correct: the clinician adjusts the segmentation if necessary
• Validate: the clinician validates the final segmentation before the measurement algorithms are applied

This hybrid AI + clinician approach aims to combine the speed of automated analysis with the reliability of clinical expertise, particularly for complex cases (poor edge delineation, higher Fitzpatrick phototypes, variable image quality).

Objectives

The study evaluates the WoundTrack 2 module along two dimensions:

• Wound surface area measurement: precision (inter-rater consistency) and accuracy (agreement with digitized planimetry)
• Tissue composition analysis: reproducibility and reliability of granulation tissue proportion measurement, a key clinical indicator of wound healing progress

Why this study?

The Pixaire-1 study, posted as a pre-print in March 2026, demonstrated the value of incorporating a clinician validation step in AI-based wound measurement. PIXAIRE-Redux represents the logical next step of this work by validating the concrete integration of the PCV workflow into an enhanced module (WoundTrack 2).

The study aims to establish the PCV workflow as a methodological standard for AI-assisted wound measurement, combining algorithmic robustness with clinician expertise. The findings will inform future deployments of automated chronic wound measurement in clinical practice and clinical research.

Keywords

Chronic wound, artificial intelligence, WoundTrack 2, PCV workflow, surface measurement, tissue composition, granulation, clinical validation, digital medical device