Automated Image-Based Wound Area Assessment in Outpatient Clinics Using Computer-Aided Methods: A Development and...

Area of Science:

• Medical imaging analysis
• Machine learning in healthcare
• Wound care technology

Background:

• Traditional wound size assessment is subjective and time-consuming.
• Variations in photo-taking distance lead to inaccurate wound measurements.
• Need for an objective, efficient wound assessment method.

Purpose of the Study:

• To develop an objective method for calculating wound size.
• To address inaccuracies caused by varying photo distances.
• To enhance consistency and efficiency in wound assessment.

Main Methods:

• Applied K-means clustering for wound segmentation based on pixel color similarity.
• Utilized a QR code as a spatial reference for accurate scaling.
• Quantified wound areas from images taken at varying distances for 40 cases.

Main Results:

• The algorithm demonstrated accuracy when tested with a standard coin.
• Paired t-tests showed no statistically significant differences between photos taken at different times (p > 0.05).
• Results indicate consistency and equivalence across multiple image captures.

Conclusions:

• The developed algorithm provides a reliable and accurate method for wound area assessment.
• The technique is interchangeable and consistently produces accurate results.
• This objective approach supports clinical decision-making and wound progression tracking.