What's your angle? Quantifying muscle pennation from diceCT scans in an open-source framework

Diffusible iodine-based contrast-enhanced computed tomography (diceCT) approaches have improved our ability to non-destructively sample soft tissue anatomy in animals. Key aspects of muscle fiber architecture, like muscle fiber length and pennation, can in theory be reconstructed from diceCT scans; however, manual segmentation is laborious and time-consuming. Several software applications have been developed to automate reconstruction of muscle fibers from diceCT data, including both open-source and proprietary options. This paper expands on the GoodFibes package, an R-language based toolkit for detecting, reconstructing, and visualizing muscle fibers from diceCT image stack datasets, to include the analysis of pennation angle in 3D. Fiber orientation is determined via principal component analysis of fiber path coordinate data, and muscle fibers may be aligned in the R environment to a tendon or line of action, permitting the direct calculation of pennation angle. I demonstrate the effectiveness of this approach using an ant mandibular muscle dataset that has been previously studied using manual digital dissection as well as both open source and proprietary approaches for muscle fiber detection and analysis. I also apply these new utilities to a comparative case study of Percid fish biting muscles. Overall, I find that this approach is effective at estimating muscle pennation angles in the ant dataset, with performance comparable to other software applications. Application of these new utilities to Percid jaw muscles reveals a pattern of divergent pennation between the adductor mandibulae (AM) pars malaris and AM pars rictostegalis in the Darter subfamily, a group of small, benthic fishes with innovative jaw anatomy and kinematics. Large, predatory Percids and small, benthic species living in streams may have different strategies for maximizing bite force proportional to body size. Overall, these additions expand the growing open-source toolkit for the analysis of diceCT data, and in particular the flexible R programming language environment that is widely used by biologists.