Quantitative evaluation of fatty change in rotator cuff muscles using the Otsu thresholding method

Area of Science:

• Orthopedics
• Radiology
• Medical Imaging Analysis

Background:

• Accurate assessment of fatty change in rotator cuff muscles is crucial for surgical planning and predicting outcomes.
• The Goutallier classification, a common grading system, is subjective and prone to inter-rater variability.
• Developing objective quantitative methods is essential for consistent fatty change assessment.

Purpose of the Study:

• To evaluate the reliability and validity of Otsu's thresholding for quantifying fatty area in rotator cuff muscles using MRI.
• To compare the performance of Otsu's thresholding method against the Goutallier classification.

Main Methods:

• Thirty-four patients with large to massive rotator cuff tears underwent MRI.
• Two surgeons independently assessed rotator cuff muscles using Goutallier classification and Otsu's thresholding via ImageJ.
• Inter-rater reliability was assessed using weighted kappa (Goutallier) and intraclass correlation coefficient (Otsu's method). Segmentation accuracy was measured using Intersection over Union (IoU).

Main Results:

• Otsu's thresholding demonstrated significantly higher inter-rater reliability (ICC >0.90) compared to Goutallier classification (kappa: 0.64-0.79).
• Segmentation accuracy (IoU) was robust (0.70-0.85), especially for supraspinatus and subscapularis muscles.
• A strong correlation existed between Goutallier grades and fatty area percentage, with significant differences across all grades (P < .05).

Conclusions:

• Otsu's thresholding provides a reliable, objective, and efficient method for quantifying rotator cuff muscle fatty change.
• This quantitative approach offers superior inter-rater reliability compared to the Goutallier classification.
• The method can serve as a standardized alternative or supplement to subjective grading systems in clinical practice.