Automatic Myonuclear Detection in Isolated Single Muscle Fibers Using Robust Ellipse Fitting and Sparse...

Area of Science:

• Biomedical Engineering
• Cell Biology
• Muscle Physiology

Background:

• Accurate myonuclear detection is crucial for understanding myonuclear domain size.
• Challenges include nuclear variability, clumping, and out-of-focus regions in z-stack images.

Purpose of the Study:

• To develop a novel automatic algorithm for robust myonuclei quantification in isolated single skeletal muscle fibers.
• To improve the accuracy and reliability of myonuclear domain size calculations.

Main Methods:

• Multi-focus image fusion for all-in-focus images.
• Heteroscedastic errors-in-variables (HEIV) regression for ellipse fitting.
• Two-stage sparse model for feature selection and classifier training.
• Modified mean-shift clustering for final nuclei detection.

Main Results:

• The algorithm was validated on 42 z-stack image sets, analyzing over 1,500 myonuclei.
• Demonstrated superior performance compared to current state-of-the-art methods.
• Achieved robust and accurate myonuclei detection despite image challenges.

Conclusions:

• The proposed automatic detection algorithm offers a significant advancement in quantifying myonuclei.
• This method enhances the reliability of myonuclear domain size analysis in skeletal muscle research.
• The algorithm provides a robust solution for challenging imaging conditions in muscle fiber studies.