Multi-method analysis for the three-dimensional reconstruction of muscle fascicles from DiceCT datasets

Area of Science:

• Comparative anatomy
• Biomechanics
• Mammalian evolution

Background:

• Muscle architecture significantly influences muscle performance, feeding ecology, and locomotion in mammals.
• Contrast-enhanced micro-CT (DiceCT) is a non-destructive imaging technique for myological data collection.
• Manual segmentation of DiceCT data is time-consuming and prone to bias.

Purpose of the Study:

• To present and validate a standardized protocol for algorithmic muscle fiber tracking using XFiber software.
• To compare the performance of XFiber against manual segmentation, an open-source algorithm (GoodFibes), and traditional dissection methods.
• To assess fascicle length and tortuosity in mammalian jaw muscles across diverse species.

Main Methods:

• A standardized protocol for algorithmic fiber tracking using XFiber (Amira suite) was developed.
• XFber performance was compared to manual segmentation, GoodFibes, and literature dissection values.
• Fascicle length and tortuosity were measured in the jaw muscles of eight mammalian species.

Main Results:

• XFber fascicle lengths closely matched gross dissection and manual segmentation, with a ~15% overestimation common to both digital methods.
• GoodFibes produced substantially longer fascicle lengths compared to other methods.
• XFber and manual segmentation showed similar tortuosity values, while GoodFibes' values were closer to manual segmentation.

Conclusions:

• The XFiber workflow offers a standardized, efficient, and reliable method for analyzing muscle architecture from DiceCT data.
• Algorithmic fiber tracking, particularly with XFiber, presents a promising alternative to manual segmentation for large-scale comparative myological studies.
• This approach minimizes time investment and inter-operator bias, enhancing the reproducibility of muscle architecture research.