Visualization and quantification of mimetic musculature via DiceCT

Introduction

Mimetic muscles—or the muscles of facial expression—are associated with social communication in visually oriented species. They can be broadly classified into three separate groups based on their location and function (muscles surrounding the external ear; muscles of the superciliary region/orbital region; and muscles of the mid-face/oral region), and all are intimately association with the skin of the face. Additionally, the platysma—which originates postcranially—inserts around the mouth and cheeks and contributes to oral expression. Over the past century, numerous anatomical reports have sought to describe the gross anatomy of these muscles in primates (e.g., Murie & Mivart, 1869; Lightoller, 1925; Sullivan & Osgood, 1925; Lightoller, 1928; Huber, 1930; Lightoller, 1934; Shibata, 1959; Seiler, 1970; Swindler & Wood, 1973; Seiler, 1977; Pellatt, 1979; Burrows & Smith, 2003; Burrows et al., 2006; Burrows, 2008; Burrows, Waller & Parr, 2009; Diogo et al., 2009; Powell et al., 2018). More recently still, analyses of variation in mimetic muscle organization between dogs and wolves have shone light on intriguing differences in communication resulting from the process of domestication (Kaminski et al., 2019). However, to date, all studies of these muscles have been limited to qualitative descriptions regarding the presence, absence, and orientation of particular muscles. This likely reflects a confluence of factors, including their small size, superficial positioning within the face, and intimate integration with the skin—each of which render the mimetic musculature difficult to individually excise and analyze using traditional gross dissection techniques. Consequently, alternative techniques for the analyses of muscle volume and configuration are necessary to more comprehensively evaluate these muscles.

In recent years, new imaging modalities such as diffusible iodine-based contrast-enhanced computed tomography (DiceCT) have emerged as a means of visualizing and quantifying myological data in situ (Gignac et al., 2016). Typically, X-ray computed tomography is reserved for the visualization of mineralized materials such as bone; however, pretreatment of specimens with a chemical stain (such as phosphomolybdic acid or, as in DiceCT, iodine) capable of binding to carbohydrates in soft tissues (Li et al., 2015) makes it possible to apply this technique to the analysis of myology. Using this technique, muscle volumes can be visualized within CT stacks, segmented, and digitally reconstructed—permitting the three-dimensional visualization of their form, and quantification of morphometric data such as surface area and volume (e.g., Cox & Jeffery, 2011; Jeffery et al., 2011; Baverstock, Jeffery & Cobb, 2013; Holliday et al., 2013; Cox & Faulkes, 2014; Gignac & Kley, 2014). Indeed, recent advances have even permitted the reconstruction of individual fascicles within whole muscle volumes (Kupczik et al., 2015; Dickinson, Stark & Kupczik, 2018; Dickinson et al., 2019; Dickinson et al., 2020; Sullivan et al., 2019).

As this methodology facilitates the visualization of subdermal tissues in situ, DiceCT circumvents the obstacles that limit our ability to quantitatively assess the relative proportions of the mimetic musculature. In so doing, it becomes possible to compare three-dimensional data on the size and configuration of these muscles across species, potentially unlocking new insights into mimetic diversity across primates within the context of communicative behavioral repertoires. Within this pilot study we demonstrate the potential of this technique for quantification of these under-studied muscles within a strepsirrhine taxon, the blue-eyed black lemur (Eulemur flavifrons). As these muscles have been qualitatively described from several closely-related species of lemur (e.g., Murie & Mivart, 1869; Lightoller, 1934; Burrows & Smith, 2003), the use of this taxon will facilitate comparison of our observations to those of these preceding studies.

Materials & Methods

Segmentation

From the reconstructed image stack (2D slice in Fig. 1; see data availability section for URL to full 3D dataset), we identified and segmented fifteen mimetic muscles: the frontalis, occipitalis, platysma, orbicularis oris, orbicularis oculi, anterior auricularis, superior and inferior posterior auricularis, superior and inferior auriculolabialis, levator labii, mentalis, mandibuloauricularis, tragicus, and depressor helicis, using Amira 6.3 (Thermo Fisher Scientific). Due to the thin nature of each muscle portion, muscles were manually segmented in multiple anatomical planes to aid interpretation and ensure that no muscle tissue was disregarded. Three-dimensional volumes (constrained smoothing, intensity = 3) were produced of each labeled muscle, from which the volume of each muscle was calculated (Table 1).

Table 1:

Volumetric measurements of key muscles of facial expression within Eulemur flavifrons.

Muscle Region	Muscle	Inferred muscle function inE. flavifrons	Volume (mm3)	Proportion of Total Mimetic Musculature (%)
Muscles of the Superciliary/Orbital Region	Frontalis	Elevation of the brow	68.86	6.4%
	Occipitalis	Posterior movement of the scalp	39.87	3.7%
	Orbicularis Oculi	Closure of the eye	329.96	30.4%
Muscles of the Midface/Mouth Region	Orbicularis Oris	Pursing of the lips	459.34	42.4%
	Superior Auriculolabialis	Elevation of the corner of the mouth	69.53	6.0%
	Inferior Auriculolabialis	Retraction of the corner of the mouth	59.94	5.9%
	Mentalis	Protrusion of the lower lip	9.65	0.9%
	Levator Labii	Elevation of the upper lip	7.03	0.6%
Muscles Surrounding the external ear	Anterior Auricularis	Elevation of the ear	3.66	0.3%
	Superior Posterior Auricularis	Flattening/retraction of the ear	14.96	1.4%
	Inferior Posterior	Flattening/retraction of the ear	8.17	0.8%
	Tragicus	Retracts/opens the ear	5.48	0.5%
	Depressor Helicis	Anterior flattening of the ear	6.18	0.6%
	Mandibuloauricularis	Anterior flattening of the ear	5.46	0.5%

DOI: 10.7717/peerj.9343/table-1

Results

Contrast-enhanced computed tomography and digital reconstruction of the facial mask of E. flavifrons permitted both three-dimensional visualization and volumetric quantification of the mimetic musculature within this specimen. In total, 14 muscle volumes were quantified, totaling 1,086.21 mm³. A visual representation of these muscles in E. flavifrons, including a superimposition of these muscles relative to the skull, is presented in Fig. 2. Volumes were measured for all muscles barring the platysma, as the infero-caudal portion of this muscle originating from the clavicle was absent from our face mask, such that any volume would be incomplete. A full summary of the mimetic muscles and their volumes is presented in Table 1.

In addition to this volumetric data, the three-dimensional visualization of these muscles also enables us to compare the configuration and organization of these muscles to previous qualitative accounts of the mimetic muscles within other species of Lemuroidea (Table 2). Several muscles closely accorded with earlier descriptions; the orbicularis oris, tragicus, anterior auricularis, platysma, and frontalis all display similar patterns of organization to existing descriptions of these muscles within Lemuroidea and Lorisoidea. Similarly, fibers from the inferior auriculolabialis arise from the superior border of platysma, following descriptions of this muscle within Otolemur by Burrows & Smith (2003) but contrasting to descriptions by Lightoller (1934). The latter noted a small gap between these muscles in Eulemur macaco and a hybrid black lemur.

A number of remaining muscles—most prominently the orbicularis oculi, superior auriculolabialis, and posterior auricularis—demonstrated significant variation from earlier descriptions of the mimetic muscles in adult strepsirrhines. Burrows & Smith (2003) report integration between fibers of the orbicularis oculi and superior auriculolabialis within Otolemur, with the o. oculi giving rise to the superior auriculolabialis. Within E. flavifrons, however, no interdigitation between fibers of these two muscles was observed when examining the muscles in digital space. We did, however, note significant integration between the superior auriculolabialis and posterior-inferior fibers of the frontalis muscle. Anterior fibers of the superior auriculolabialis both integrate with fibers from the posterior frontalis and, in some regions, overlap such that these fibers sit immediately deep to overlying fascicles from the frontalis. This integration is not reported within Otolemur (Burrows & Smith, 2003) nor Eulemur macaco and the hybrid black lemur described by Lightoller (1934).

Finally, we note variation in the configuration of the posterior auricularis from descriptions of this muscle by both (Lightoller, 1934) and Burrows & Smith (2003). Within Eulemur macaco and the hybrid black lemur, this muscle is described as comprising three bellies, while Burrows & Smith (2003) describes two distinct bellies of equal size in Otolemur. The organization of this muscle in E. flavifrons is similar to that of Otolemur in comprising two bellies; however, within our specimen the superior belly possesses a volume almost twice as large as the inferior belly (14.96 vs. 8.17 mm³).

Discussion

Digital reconstruction of the facial mask of E. flavifrons makes it possible to report, for the first time, quantitative data on the size of these highly-integrated muscles (Table 1). Our visualizations also enable us to compare the configuration and organization of these muscles to previous qualitative accounts of the mimetic muscles within other species of Lemuroidea. As described above, several muscles closely accorded with earlier descriptions. Additionally, a minor deviation from one earlier description was observed within the inferior auriculolabialis.

A number of remaining muscles demonstrated more significant variation from earlier descriptions. Burrows & Smith (2003) report integration between fibers of the orbicularis oculi and superior auriculolabialis within Otolemur, with the O. oculi giving rise to the superior auriculolabialis. Within E. flavifrons, however, no interdigitation between fibers of these two muscles was observed when examining the muscles in digital space. We did, however, note significant integration between the superior auriculolabialis and posterior-inferior fibers of the frontalis muscle. Anterior fibers of the superior auriculolabialis both integrate with fibers from the posterior frontalis and, in some regions, overlap such that these fibers sit immediately deep to overlying fascicles from the frontalis. This integration is not reported within Otolemur (Burrows & Smith, 2003) nor Eulemur macaco and the hybrid black lemur described by Lightoller (1934). Finally, we note variation in the configuration of the posterior auricularis from descriptions of this muscle by both Lightoller (1934) and Burrows & Smith (2003). Within Eulemur macaco and the hybrid black lemur, this muscle is described as comprising three bellies, while Burrows & Smith (2003) describes two distinct bellies of equal size in Otolemur. The organization of this muscle in E. flavifrons is similar to that of Otolemur in comprising two bellies; however, within our specimen the superior belly possesses a volume almost twice as large as the inferior belly (14.96 vs. 8.17 mm³).

By combining the reverse-dissection technique with a high-resolution DiceCT protocol, we successfully establish herein a novel method for the three-dimensional visualization and quantification of the mimetic musculature. This study provides further evidence of the value of contrast-enhanced tomographic techniques in permitting the quantification of skeletal muscle volumes, following earlier studies (e.g., Cox & Jeffery, 2011; Baverstock, Jeffery & Cobb, 2013; Lautenschlager, 2013; Cox & Faulkes, 2014; Dickinson, Stark & Kupczik, 2018). Indeed, this protocol enabled the quantification of muscles with volumes as small as 4–10 mm³. However, it should be noted that not all portions of the face could be visually resolved; specifically, the nasal/rostral region presented small muscles which were highly integrated with connective tissues in an inconsistent manner, such that anatomical boundaries between muscle bodies could not be confidently discerned, and individual fascicles could typically not be resolved. It is therefore possible that sub-region scanning of this anatomy at even higher (<20 µm) resolutions may be necessary to accurately quantify this musculature.

Though these initial results appear promising, several caveats to the present study should be noted. Firstly, the inherent intra-specific variation in the organization of the mimetic musculature (as demonstrated in humans; e.g., (Watanabe, 2016)) precludes confident interpretation of evolutionary trends from single-specimen samples. Additionally, it is important to consider the potential for volumetric modification of muscles during the preparation process. Indeed, excision, the dessication of extraneous tissues, and staining could all potentially impact the final volumes by inducing shrinkage of superficially exposed tissues. Therefore, while the current study applies this protocol to an isolated facemask, future studies of mimetic muscle organization may wish to explore the anatomy of these muscles in situ. Although this method would likely necessitate a longer staining time, the mimetic muscles can be clearly seen in some DiceCT specimens that have been prepared for analysis of other regions (e.g., the focal specimen from (Dickinson et al., 2020)). In situ quantification would also enable the visualization of mimetic muscle tissues in direct association with the underlying morphology of the face, as well as minimizing the potential for muscle deformation and shrinkage associated with the excision and staining process. Though this effect appeared relatively minimal for the muscles analyzed herein, the extreme borders of several muscles displayed a slight curvature interpreted to reflect an unnatural curling of the muscle tissue into itself. In situ analyses in which surrounding tissues can stabilize these muscles may therefore yield even more precise data on muscle organization within the face. This would in turn allow, for the first time, the three-dimensional visualization of mimetic muscles from their origins to their insertions.

Conclusions

The application of DiceCT permits both the visualization of the mimetic musculature in three-dimensional space and, for the first time, the quantification of muscle volumes for these small and highly-integrated tissues. These novel data further demonstrate the potential for contrast-enhanced tomographic techniques in enhancing our anatomical understanding of small and obscured structures. Future studies may apply an adapted protocol to visualize these tissues in situ within other taxa, and provide more spatial context for these muscles in relation to other tissues of the face.