An efficient framework for estimation of muscle fiber orientation using ultrasonography

Area of Science:

• Medical Imaging
• Biomedical Engineering
• Musculoskeletal Ultrasound

Background:

• Muscle fiber orientation (MFO) is crucial for understanding musculoskeletal function.
• Traditional manual MFO estimation from sonograms is time-consuming.
• Existing automated methods for MFO estimation are computationally intensive due to voting procedures.

Purpose of the Study:

• To develop a novel and efficient framework for estimating MFO in sonograms.
• To improve the computational efficiency of automated MFO estimation.
• To maintain or enhance the accuracy of MFO measurements.

Main Methods:

• Utilized Multi-scale Vessel Enhancement Filtering (MVEF) to enhance muscle fascicles in sonograms.
• Applied binarization to the enhanced images followed by detection of line-shaped patterns.
• Implemented direct orientation computation for elongated regions and optional Hough voting for other cases, reducing dependency on voting.

Main Results:

• The proposed method demonstrated strong agreement with manual MFO measurements.
• Achieved significantly higher efficiency compared to the Re-voting Hough Transform (RVHT) algorithm.
• Experimental validation across diverse sonogram datasets confirmed performance.

Conclusions:

• The developed framework accelerates MFO estimation by minimizing reliance on voting procedures.
• Accuracy is preserved, indicating a more efficient alternative to existing methods.
• The study successfully enhances the speed of orientation estimation algorithms for sonographic analysis.