Modelling of chronic wound healing dynamics

Area of Science:

• Wound healing research
• Mathematical modeling in medicine
• Biostatistics

Background:

• Tracking chronic wound area over time provides insights into healing.
• Existing two-parameter models (linear, exponential) often fail to capture healing delays.
• Advanced models are needed for accurate chronic wound assessment.

Purpose of the Study:

• To evaluate the efficacy of two-, three-, and four-parameter mathematical models in describing chronic wound healing.
• To compare model performance using quantitative and qualitative criteria.
• To identify the most suitable model for clinical application in chronic wound management.

Main Methods:

• Weekly measurements of 226 chronic wounds across various etiologies.
• Assessment of two-, three-, and four-parameter models using goodness of fit, missing data handling, and prediction error.
• Qualitative evaluation based on parameter count and biophysical meaning.

Main Results:

• Three- and four-parameter models showed significantly better goodness of fit (median 0.937-0.958) than two-parameter models (median 0.821-0.883).
• Prediction errors were lower for three-parameter models (64-128) compared to two-parameter (111-746) and four-parameter (238-407) models.
• The delayed exponential model was identified as the most appropriate three-parameter model.

Conclusions:

• Three- and four-parameter models provide a more accurate representation of chronic wound healing dynamics.
• The delayed exponential model demonstrates good prediction capability, aiding clinical decision-making.
• This model can assist physicians in selecting optimal treatments after an initial three-week observation period.