Evaluation of a Semi-Automated Wound-Halving Algorithm for Split-Wound Design Studies: A Step towards Enhanced...

Area of Science:

• Biomedical Engineering
• Wound Care Research

Background:

• Chronic leg ulcers pose a significant global healthcare challenge.
• Accurate wound measurement is crucial for evaluating treatment efficacy.
• Existing methods may lack objectivity and efficiency.

Purpose of the Study:

• To validate a novel "split-wound design" approach using objective measures.
• To introduce and evaluate a semi-automated wound measurement algorithm.
• To compare algorithm-based measurements against manual assessments.

Main Methods:

• Developed a semi-automated algorithm to objectively determine a wound halving line.
• Calculated the halving line by splitting the wound's bounding box along its longest axis.
• Compared algorithm-generated halving lines with manual operator assessments and ground truth data.

Main Results:

• The algorithm demonstrated significantly higher accuracy than manual measurements.
• The algorithm's median absolute deviation (MAD) from ground truth was 0.3%, compared to 2-3% for manual methods (p < 0.001).
• This indicates superior reliability and precision of the automated approach.

Conclusions:

• The developed wound-halving algorithm is reliable and suitable for wound studies.
• Combining the algorithm with the split-wound design offers benefits like reduced patient recruitment, faster planning, and cost savings.
• This innovative approach accelerates evidence generation in wound care research and practice.