Pressure ulcer image segmentation technique through synthetic frequencies generation and contrast variation using...

Area of Science:

• Medical Imaging
• Computational Biology
• Digital Image Processing

Background:

• Pressure ulcers are a growing concern due to high treatment costs and reduced patient quality of life, particularly affecting the elderly and disabled.
• Accurate diagnosis and treatment of these chronic wounds are time-consuming, often taking weeks or months.
• Non-invasive image processing techniques offer a promising approach to analyze ulcers and assist in diagnosis.

Purpose of the Study:

• To develop and evaluate a novel image segmentation technique for pressure ulcers.
• To improve the accuracy and efficiency of pressure ulcer analysis for better diagnosis and treatment planning.

Main Methods:

• A new image segmentation technique based on contrast changes using synthetic frequencies derived from pixel grayscale values.
• Utilizes toroidal geometry to decompose images into contrast levels and Otsu's thresholding for binarization.
• Employs morphological operations to refine segmented regions.

Main Results:

• The technique was evaluated on a database of 51 pressure ulcer images.
• Achieved an average correlation of 0.89 ±0.1 with actual segments using digital image correlation.
• Demonstrated a computational time of 9.04 seconds.

Conclusions:

• The proposed methodology provides superior segmentation results compared to benchmarked algorithms.
• Offers a faster computational time and does not require initial conditions for segmentation.
• Aids in the diagnosis and treatment of pressure ulcers through improved image analysis.