Despite the growing clinical relevance of myosteatosis, the biological mechanisms underlying increases in myosteatosis with aging in humans remain considerably unknown. Current theories are largely based on animal models or controlled experimental studies in small samples of humans. Some propose that increased myosteatosis may result from increased fatty acid transport, uptake and storage, and reduced fatty acid oxidation (Kim et al., 2000; Kelley and Goodpaster, 2001). Other experiments suggest that macrophage infiltration inhibits adipocyte differentiation, and leads to adipocyte hypertrophy, altered cytokines secretion, and increased fat accumulation in muscle and other non-adipose tissues (Hegarty et al., 2002). It is also thought that an excess lipid availability and high-fat diet may reduce the mitochondrial respiratory chain components in muscle. Impaired skeletal muscle mitochondrial function, which directs fatty acids toward storage as opposed to oxidation, may also contribute to aging-related increases in skeletal muscle fat accumulation (Roden, 2005; Mathieu et al., 2010). Another potential mechanism for these alterations involves the possible age-related change in activation, proliferation, and differentiation of quiescent skeletal muscle precursor “stem” cells (i.e., satellite cells) into adipocytes in response to a range of stimuli that occurs with aging. Inadequate storage of excess fat in subcutaneous fat depots during excess lipid availability has also been identified as a potential cause of the increased myosteatosis. The subcutaneous fat depot may act as a metabolic “sink,” which can accommodate excess lipids and thus prevent the flow of lipid to vital organs. A better understanding and identification of the biological factors that regulate myosteatosis with aging, including the evidence of senescent cells and cultured cells developing into preadipocytes and fat cells (Tchkonia et al., 2010), is imperative for developing new strategies for promoting metabolic and musculoskeletal health. Such biomarkers could become targets for novel therapies to preserve a more functionally and metabolically “healthy” skeletal muscle with aging (Mathieu et al., 2010; Tchernof and Després, 2013; Smith, 2015).

Though the specific mechanisms underlying the detrimental effects of myosteatosis on local and systemic muscle metabolism remain unknown, several hypotheses have been proposed that support this relationship (Albu et al., 2005; Gallagher et al., 2005; Durheim et al., 2008; Lee et al., 2009; Vettor et al., 2009; Kovalik et al., 2011; Trayhurn et al., 2011). It is theorized that the close proximity of fat to the muscle fiber may impair the local muscle environment through elevated levels of local or systemic pro-inflammatory cytokines (Vettor et al., 2009; Zoico et al., 2010) or throughout the impaired myokine secretion (Brandt and Pedersen, 2010). Increased myosteatosis may also impair nutritive blood flow to muscle and thus impair insulin action and insulin diffusion capacity (Lee et al., 2009). Additionally, recent findings indicate that an increased number of inter-muscular adipocytes have a direct local impact on skeletal myocyte metabolism, causing enhanced myotube mRNA expression of genes involved in oxidative metabolism (Lowell and Shulman, 2005; Kovalik et al., 2011). This study suggested that even a relatively small amount of intermuscular fat may be sufficient to drive myotubes into lipid oxidation and affect skeletal muscle metabolism. However, these effects depended on the metabolic state of the system. It is also hypothesized that the increasing rate of lipolysis within skeletal muscle results in an elevated concentration of glucose within the skeletal muscle (Yim et al., 2007; Turcotte and Fisher, 2008). From animal models, it has been shown that insulin resistance observed in old high-fat diet fed rats is independently related to myosteatosis, and that it can be induced by high ceramide synthesis (Tardif et al., 2014), stimulated by adipocyte-derived cytokines (including tumor necrosis factor-α; Summers and Nelson, 2005; Tardif et al., 2014) and consequent decrease in the ability to upregulate muscle insulin pathways.

In addition to the possible effects of myosteatosis on muscle and mobility function through changes in local muscle metabolism, myosteatosis may also be harmful to muscle and mobility function due to mechanical changes in muscle that occur in the presence of myosteatosis, leading to changes in muscle fiber orientation (Yoshida et al., 2012a). Along with the loss of elasticity in rotator cuff muscles, it has been hypothesized that excess myosteatosis leads to an alteration in contractile fiber pennation angle, hence resulting in an unfavorable mechanical angle and a concomitant reduction in force production (Meyer et al., 2004; Gerber et al., 2007). To our best knowledge, no studies have examined the effect of myosteatosis on elasticity or of pennation angle in locomotor muscles. Finally, pro-inflammatory cytokines (Vettor et al., 2009) secreted by adipose tissue in the skeletal muscle microenvironment may lead to proteolysis and muscle catabolism (Ebisui et al., 1995; Hegarty et al., 2002; Zoico et al., 2010). Figure 1 summarizes the potential mechanisms underlying the effects of myosteatosis.

Image-based methods of assessing myosteatosis are applied primarily in research settings, and are not yet in routine clinical use. Based on these facts, there is a need to establish an objective, valid, reliable, and standardized measure of quality derived from the composition of the skeletal muscle. MRI of a cross section of the thigh of young and old males and females matched for weight and height demonstrate that fat infiltration is not an issue of being overweight, but a sign of important changes with aging (Petrella et al., 2007). MRI is an excellent non-invasive technique for measuring myosteatosis, providing high-quality images. However, the cost is high, and traditional MRI does not typically allow quantification of the fat content of the muscle. Recent MRI work is aiming to assess myosteatosis using a method for fat/water separation. Chemical shift-based MRI has demonstrated an advantage over single voxel spectroscopy in readily determining tissue heterogeneity within muscle fat fractions in both normal and dystrophic muscles (Triplett et al., 2014).

Other imaging techniques also facilitate the non-invasive evaluation of myosteatosis. Computed tomography (CT) has thus far probably been the most utilized as a research tool to investigate myosteatosis. It is basically assessed indirectly using muscle attenuation calculated CT imaging scan, leading to close correlation with direct measurements of muscle lipid content (Machann et al., 2003; Larson-Meyer et al., 2006). Muscle radiation attenuation is a radiological characteristic that can be measured (Goodpaster et al., 2000; Goodpaster, 2002). Any given skeletal muscle displays radiation attenuation between −190 and +150 Hounsfield units (HU), with a prominent peak near +50 HU. When muscle cross-sectional area and attenuation are reported, the most common practice is to use pre-established HU ranges to define intermuscular fat (usually −190 to −30 HU) and muscle tissue (usually −29 HU to 150 HU; Aubrey et al., 2014). Although radiation exposure is a disadvantage, compared to CT scans of the trunk, the effective dose to the extremities is typically very small. CT also cannot isolate individual muscle groups. New dual or multispectral CT imaging methods may offer improved tissue discrimination in future studies.

Standard MRI, CT, and DXA imaging are techniques frequently utilized in large clinical trials, but one of their major limitations is that they cannot directly measure the lipid content or detect the location of fat storage within myocytes (intramyocellular fat) or lipid droplets surrounding myocytes (extramyocellular fat, which are present in very small quantities; Goodpaster, 2002; Machann et al., 2003; Karampatos et al., 2016). Due to its biochemical specificity, the use of magnetic resonance spectroscopy (MRS) is a valuable approach in measuring intramyocellular lipids. However, one of its limitations is high cost. In addition, while MRS can be done on many of the newer MRI machines, it requires more advanced work to set up and is unlikely to be available at most clinical facilities. It should also be noted that ultrasonography equipment is an emerging option for muscle tissue imaging. Ultrasound is commonly available at clinical centers, portable, and much cheaper than other imaging methods. However, like other common imaging methods, ultrasound has limitations in distinguishing between intramyocellular and extramyocellular fat.

While we are primarily relying on imaging techniques at this time, there is a pressing need for developing a rapid, easy to use, portable and cost-effective method for assessing myosteatosis in clinical settings. Fat infiltration in skeletal muscles increases with aging and influences muscle's metabolic and contractile function. A growing area of research on myosteatosis would benefit from a standardized approach to quantify myosteatosis across research studies. In addition, it is important to consider the population studied and the questions posed. More basic science and epidemiologic studies will be needed to determine the mechanisms responsible for increased myosteatosis with aging, as well as mechanisms underlying the effects of myosteatosis.

The functional impact of intermuscular fat has been demonstrated by several studies in which higher intermuscular fat was associated with poorer muscle function and strength (Goodpaster et al., 2001; Visser et al., 2002; Manini et al., 2007; Hilton et al., 2008; Delmonico et al., 2009). This appears to be related to a functional contractile hindrance within the muscle (Cesari et al., 2006; Lauretani et al., 2006; Schrager et al., 2007). A closer examination of muscle structure reveals how intramuscular fat may impede muscle function. With aging, the motor unit array may become compressed due to a decrease in both size and number of muscle fibers, particularly in type II fibers (Tomonaga, 1977; Larsson and Karlsson, 1978; Lexell et al., 1988). In fact, autopsy studies (Lexell et al., 1983) reveal 25% fewer muscle fibers in the medial vastus lateralis of older adults (72 ± 1 years) in comparison to younger individuals (30 ± 6 years). Moreover, aged skeletal muscle shows a variety of changes within the extracellular matrix, including collagen accumulation and altered elasticity (Kragstrup et al., 2011). The compressed motor unit array resulting from myosteatosis and non-contractile tissue accumulation may impede muscle activation and force production in older adults (Yoshida et al., 2012b).

In addition to muscle fiber size and number, biopsy studies also show changes in other qualitative characteristics of aged skeletal muscle such as changing fiber type distribution, in which the percentage and area of type II fibers in the vastus lateralis are reduced (Larsson and Karlsson, 1978). Accompanying changes in fiber type distribution within skeletal muscle are alterations in metabolism characterized by decreases in oxidative enzyme activity and muscle capillarization (Rogers and Evans, 1993). Examination of muscle biopsies of less functional or “clinically weak” muscle compared to “normal aged muscle” reveals qualitative differences in fiber packing, shape, intramuscular lipid distribution, and connective tissue (Fragala et al., 2014). Yet, despite the detailed characteristics of the individual muscle fibers (or cells) provided in muscle biopsy analysis, routine muscle biopsy procedures using the Bergstrom needle biopsy technique (Bergstrom, 1975) are invasive procedures not suitable for routine muscle quality assessment. Recent developments in fine needle biopsy techniques show promise for muscle histology assessment (Townsend et al., 2016) with far less pain and discomfort (Hayot et al., 2005).

Cross-sectional muscle assessment yields important information about muscle function, as both individual muscle fiber diameter and cross sectional diameter of whole muscle are associated with muscle function as strength (Verdijk et al., 2010). Yet whole muscle contracts with the longitudinal shortening of muscle fibers by the sliding on the actin and myosin within the sarcomere units. Thus, assessment of both longitudinal and cross-sectional features of muscle architecture each provide important insights into muscle functioning. Ultrasonography enables the valid and reliable assessment of a variety of muscle architectural characteristics, such as anatomical cross-sectional area, muscle thickness, pennation angle, fascicle length, and echogenicity measured using grayscale analyses (i.e., a measure related to muscle density and myosteatosis; Cadore et al., 2012; Scott et al., 2012; Cartwright et al., 2013). Moreover, some evidence exists to show that local muscle hypertrophy may be detected on ultrasound before changes in DXA are apparent (Scanlon et al., 2014).

The muscle quality domain of force production is related to the architectural characteristics of skeletal muscle (Gans and de Vree, 1987; Gans and Gaunt, 1991; Fukunaga et al., 1997; Lieber and Friden, 2000), including muscle fiber length and arrangement in relation to the direction of force produced by the whole muscle (Gans and de Vree, 1987). Consequently, both cross sectional and longitudinal orientation measures of skeletal muscle have value in the evaluation of the size-strength relationship and the existence of age-related differences in muscle strength per size (Akagi et al., 2009). Longitudinal assessment by electron microscopic analysis of human skeletal muscle reveal sarcomeric changes in myofibrillar disorder, Z-line streaming, and dilatation in aged skeletal muscle (Scelsi et al., 1980). In consideration of such factors, architectural characteristics can be combined into a composite measure, often referred to as physiological cross-sectional area (PSCA)—a measure that is related to both leg extension strength and change in leg extension strength in older adults (Scanlon et al., 2014).

Muscle mass is strongly coupled with muscle strength during the course of maturation through young adulthood. However, it has been observed that the relationship between muscle mass and strength decreases in magnitude as one ages, and that the age-related loss of muscle strength may occur more rapidly than the loss of LBM (Landers et al., 2001; Ferrucci et al., 2011). The measurement of muscle strength remains a vital outcome within the realm of geriatric assessment. Nevertheless, the isolated detection of low muscle strength cannot be used for staging or diagnosis, since there are varied causes of myogenic and neurogenic weakness that require additional medical follow up for the differential evaluation. Moreover, there is some evidence that diminished muscle tissue composition may precede losses in muscle mass as a contributing factor to the age-related decline in strength (Visser et al., 2005; Goodpaster et al., 2006; Watanabe et al., 2013; Ismail et al., 2015).

There are varied approaches to assessing the microscopic and gross morphological features of skeletal muscle tissue that may affect muscle quality. Muscle architecture consists of the macroscopic morphology of skeletal muscle based on the muscle shape and its fiber arrangement relative to the axis of force generation (Anneriet et al., 2007; Lieber, 2010). Whereas, muscle ultrastructure (also called, muscle microstructure) is the microscopic characterization of muscle tissue and its specialized cellular components and organization. This approach to myology dates back to Van Leeuwenhoek's light microscope studies in 1712 and was more fully elucidated in the seminal publication of the “Skeletal Muscle” Section of the Handbook of Physiology by the American Physiological Society in 1983 (Peachey et al., 1983). The study of muscle ultrastructure traditionally includes cytology, histology, surface morphology, and other anatomical approaches (Lieber, 2010; Eisenberg, 2011).

Regarding body composition assessment, multiple models exist which range from the molecular and cellular level to the tissue/organ and whole body level. However, all estimates of body composition are limited by a lack of true direct measurement (Heymsfield et al., 2015). Composition estimates are largely based on biochemical or tissue density qualities, with imaging methods typically incorporating a tissue/organ level of measurement (i.e., skeletal muscle tissue, adipose tissue, osseous tissue, visceral organs, and other tissues) or a whole-body level of measurement (i.e., head, trunk, and appendages) based on fat mass, fat-free mass, and bone mass. The use of imaging modalities as a non-invasive means of body and tissue composition assessment has a long history of usage in medical facilities. Although DXA, MRI, and CT are deemed acceptable tools for this purpose, significant barriers limit their incorporation into the sarcopenia staging process. DXA is perhaps the most widely accepted imaging modality for the estimate for LBM (expressed as total LBM or aLM/ht²) for sarcopenia staging, but it does not yield useful information concerning muscle quality based on adverse changes in muscle tissue composition. Both MRI and CT do provide estimates of muscle tissue composition, but CT is limited by its relative high ionizing radiation exposure, and MRI use is often beset by concerns related to access and cost. Standard segmental bioimpedance analysis (BIA) or refinements such as multi-frequency electrical impedance analysis could prove to be a non-imaging modality capable of characterizing elements of muscle quality. The phase angle derived from BIA has been associated with diminished strength in older adults and serial measures may partially reflect age-related changes in muscle tissue (Heymsfield et al., 2015). Both BIA and diagnostic ultrasound have comparative advantages, with the former having a low cost suitable for field use, and the latter having the additional ability to capture morphological features of skeletal muscle. Diagnostic ultrasound has been proposed as an alternative imaging modality for the use of sarcopenia screening and staging due to its portability, relatively low cost, and potential ability to provide proxy estimates of muscle tissue composition (Reimers et al., 1993; Sipilä and Suominen, 1993; Harris-Love et al., 2014).

Recent results from Cawthon and associates (Cawthon et al., 2015) involving the Osteoporotic Fractures in Men (MrOS) cohort were consistent with previously reported findings regarding the association of LBM with reported falls, limitations in mobility, and mortality. However, these investigators noted that staging individuals for sarcopenia using the major LBM cut off values only served to decrease the estimated risk of an adverse outcome in people within the cohort that had an event. The predictive value of LBM concerning morbidity and mortality in these community-dwelling older men did not exceed the predictive value of age alone. In contrast, findings from other studies involving large cohorts of community-dwelling older adults have indicated that muscle strength and muscle quality (based on ratios of strength and regional LBM, or tissue density algorithms involving Hounsfield units derived from CT scanning), but not muscle size or LBM, portend increased risk of hospitalization and incident disability (Cawthon et al., 2009; Hairi et al., 2010). Estimates of muscle quality based on CT scanning estimates of tissue density have been shown to be significantly associated with ultrasound echogenicity in samples of older women (Sipilä and Suominen, 1993). Moreover, preliminary evidence supports the observation that high echogenicity values obtained from anterior quadriceps ultrasound scans are associated with decrements in muscle force independent of muscle mass levels in older adults (Watanabe et al., 2013; Ismail et al., 2015).

A recent study featuring an assessment of muscle mass via DXA (e.g., aLM/ht²) and echogenicity as a proxy estimate of muscle tissue composition revealed that estimates of muscle quality via sonographic imaging were independently associated with relative grip strength, whereas LBM was not (Ismail et al., 2015). Upon accounting for both age and body fat in the relationship between echogenicity (as measured using grayscale histogram analysis; Harris-Love et al., 2016b) and strength, both echogenicity and age were identified as significant factors. However, echogenicity superseded age as a factor in the preliminary model (r_xyz = −0.52 vs. r_xyz = −0.38). Indeed, the study participants with higher levels of BMI and LBM (LBM > 6.75 kg/m²) exhibited lower levels of relative grip strength in comparison to those with lower levels of BMI and LBM with Class I sarcopenia (LBM = 5.76–6.75 kg/m²; Figure 3; Janssen et al., 2004a; Ismail et al., 2015). The collective observations of investigators (Visser et al., 2005; Goodpaster et al., 2006; Watanabe et al., 2013; Ismail et al., 2015) using imaging modalities to assess skeletal muscle in older adults raise the question, “Should an approach to muscle quality involving tissue composition be introduced as an independent criterion measure within the staging concept for pre-sarcopenia?” In order to address this question, additional studies should validate these reported observations in larger samples. These studies will allow better understanding of whether estimates of muscle tissue composition can be utilized as predictors of important clinical outcomes in older adults.

Key factors that affect quantitative sonography procedures include the parameters associated with image acquisition, the methods used during image processing and analysis, the psychomotor skills of the examiner, and the wide array of technological capabilities among commercially available ultrasound machines. The selection of superficial muscle groups to determine sonographic estimates of LBM and muscle tissue composition confers some methodological advantages. Superficial muscle groups are readily identified via gross anatomical landmarks during the ultrasound examination. Also, their use minimizes excessive attenuation of the sound waves due to either large amounts of adipose tissue or the greater depth of penetration required for deeper tissues (Harris-Love et al., 2014). Typically, this approach to musculoskeletal imaging is aided by the use of high frequency linear ultrasound transducers which have a larger attenuation coefficient and may yield higher image resolution of the targeted muscle groups. The use of anterior axial and appendicular muscle groups have been proposed for ease of access during the ultrasound examination for sarcopenia staging (Ismail et al., 2015), but other approaches have been effectively used to derive estimates of LBM (Abe et al., 1994). Additional image acquisition parameters beyond the selected scanning sites may also foster imaging consistency. Many muscle groups with architectural features such a fusiform shape or convergent/triangular structure near the tendinous insertion may benefit from scanning in the longitudinal view. Previous findings from Zaidman and associates (Zaidman et al., 2008) indicated that longitudinal view images may be more reliable than the use of transverse view measurements when calculating calibrated muscle backscatter values from elbow flexor scans (intraclass correlation coefficients, ICCs, of 0.93–0.95 vs. 0.73–0.82). The method used to identify the region of interest (ROI) within the field of view may also impact estimates of muscle tissue composition using echogenicity values. While both near field and full muscle thickness ROI methods have been employed to characterize aberrant changes in echogenicity, the heterogeneity of age-related alterations in muscle tissue may merit the selection of the full muscle thickness ROI during image analysis (Watanabe et al., 2013; Harris-Love et al., 2016b).

Quantitative approaches to diagnostic musculoskeletal ultrasound have been shown to be reliable for the measurement of axial and appendicular muscle groups under controlled conditions (O'Sullivan et al., 2007; Sions et al., 2014). Moreover, investigators using quantitative sonography have exhibited good reliability for morphometry and for measures of echogenicity using grayscale histogram analysis and calibrated backscatter calculations (Zaidman et al., 2012, 2014). Nevertheless, the operator dependency of ultrasound measurements still exceeds the influence of the examiner in standard DXA and CT assessments of LBM. Although experience and training certainly increases scanning and image acquisition consistency between examiners, new modes of enhanced sonography have been proposed to promote reliability across examiners during serial imaging (Gilbertson and Anthony, 2013; Harris-Love et al., 2016a). Using a standard ultrasound transducer augmented with axial force sensors, examiners with varied levels of experience have demonstrated consistent material thickness measures over a range of applied scanning forces (Harris-Love et al., 2016a). Examiners with experience classifications of novice (1 month), intermediate (1 year), and experienced (>10 years) attained a high degree of interrater reliability (ICC_{2, k} = 0.97, p < 0.001; coefficient of variation = 1.5–2.9%) upon measuring a muscle mimetic ultrasound phantom utilizing up to 10 N of applied force. This demonstrated ability to obtain consistent image measurements at relatively high forces may have clinical implications given that the observed applied forces by sonographers may range from 5 to 14 N during abdominal examinations (Gilbertson and Anthony, 2013).

The development of enhanced force control scanning methods may continue to evolve and provide a viable alternative for sonographic imaging in varied clinical settings under realistic practice conditions. The technological advances of sonography are developing at a rapid pace in comparison to the contemporary innovations associated with DXA imaging. The continual improvement in sonography technology, coupled with the ongoing need to rectify image properties with tissue properties across different ultrasound devices, will be the source of opportunities and challenges in the quest for alternative imaging assessments of muscle quality.

Variations in ultrasound probe contact force lead to variations in ultrasound images, making the images difficult to directly and quantitatively compare with each other. The contact force required to obtain an ultrasound image deforms the underlying soft tissue. Conventionally, the probe-patient contact force is controlled qualitatively by the human operator. A force-measurement platform (Figure 4) has been developed, which can simultaneously image and measure precise, operator-applied, force. During an ultrasound exam, it was demonstrated that contact forces exerted by the operator can vary, up to 50% over 30 s, resulting in images that are acquired at different levels of tissue deformation (Gilbertson and Anthony, 2013). This force measurement platform has been used to explore the impact of preload the on elastic property estimates of Young's modulus in tissue and to characterize the average preload applied during abdominal sonography (Dhyani et al., 2014); these values range between 4.4 and 10 N. This preliminary analysis reveals trends in applied force vs. years of experience and BMI of the patient. Shown in Figure 5 are ultrasound images of the biceps from a healthy subject, at four different applied forces. The muscle thickness as measured from the bone to the subcutaneous fat-muscle separation layer is highly dependent on force. Variable compression impacts the quantities extracted from ultrasound image analysis. Physical parameters such as elasticity are biased along the tissue's non-linear stress strain curve. Image-based parameters such as average grayscale level or and image texture are directly altered under different levels of tissue compression.

Duchenne muscular dystrophy (DMD) is characterized by progressive disability leading to death and it remains one of the most common and devastating neuromuscular disorders of childhood (Bushby et al., 2010). It is caused by a genetic mutation that generates a complex sequence of events in muscle cells, which eventually undergo fibrosis and are replaced by adipose and connective tissue. Although a variety of promising new treatment strategies are in development, outcome measures for clinical trials remain limited for the most part to a set of functional measures, such as the 6-min walk test (Enright, 2003). While clearly useful, such measures are impacted by unrelated factors, such as mood and effort, and have limited repeatability. To address this and other limitations, magnetic resonance imaging is now being investigated as a surrogate measure (Willcocks et al., 2014). However, more easily applicable, cost-effective, office-based surrogate measures that provide high repeatability and sensitivity while still correlating strongly with disease status would find wider use. The information embedded in acquisition-state-controlled ultrasound images may provide convenient, non-invasive, clinically meaningful markers of muscle disease progression that surpass the functional measures currently in use.

Clinically meaningful markers of muscle disease progression in DMD are currently being developed. The potential for force-correlated ultrasound imaging to enable automated classification of muscle tissues structures in individuals with DMD has been shown. Image-based quantities, such as variance maps (Koppaka et al., 2014a), and image texture (Koppaka et al., 2014b), have been used in classification schemes. Automated classification was found to be improved when the biomarkers extracted (from the quadriceps, biceps, forearm, tibialis anterior, and medial gastrocnemius) were tagged with force information. Similar methods may be used to promote ultrasound as a quantitative non-invasive tool for not only classifying disease severity, but also for tracking progression over time.

Ultrasound elastography, whether strain elastography or shear wave elastography (SWE), also suffers from operator-dependent and acquisition state variation discussed above. Ultrasound elastography limitations have been identified by scientists and clinicians in the context of technique performance, clinical utility, and work-flow. The most frequently acknowledged limitation is repeatability and reproducibility. A few confounding factors introduce bias and variation in stiffness measurement, including operator dependence, concerning both strain elastography and SWE, and cross-vendor system dependence of SWE caused by different shear wave excitation mechanisms, tissue mechanical models, and signal processing involved in stiffness reconstruction. Before ultrasound elastography can be fully adopted as a routinely used clinical tool for diagnosis and therapy monitoring, these technical and clinical challenges must be overcome by the entire industry.

Soft tissue is mechanically non-linear, meaning that elastography measurement varies with the operator-applied transducer force (i.e., preload). Preload-induced variation may lead to inaccurate diagnosis and poses challenges to establishing measurement standardization and clinical guidelines. Even with SWE, which is considered more operator-independent than strain elastography, stiffness measurement on the same tissue can vary significantly. The results in Figure 4 show significant variation in shear wave elastography estimates of tissue elasticity as a function of preload differences typical of clinical sonography. The varying preload conditions depicted, 1–18 N, are typical of those seen across a range of operators in routine sonography (Dhyani et al., 2014) and the resultant change in estimated tissue Young's modulus ranged from 21.1 to 64.1 kPa in the vastus medialis. This variation is explained by the observation that different bias compression levels pre-strain the tissue to different operating points along the tissue's non-linear stress-strain response.

Knowing and controlling the position and orientation of an ultrasound probe on the body is also a central issue to reducing variability over serial examinations. Local skin appearance can be used to help relocate a probe on the body. This is accomplished by attaching a small camera to an ultrasound probe and capturing high magnification images of the skin at the same time as the ultrasound images.

Patterns of skin features are unique when viewing a sufficiently large area of skin and include micro reliefs, follicles, moles, and melanin variation (Figure 6). This uniqueness allows skin to be used as a relative-motion or absolute-location encoding system for the body. Based on this concept, a freehand ultrasound platform has been developed which localizes an ultrasound probe in 6-degree-of-freedom (DOF) with respect to the patient (and not room coordinates) by tracking natural skin features (i.e., fiducial-free) with a small camera mounted to the ultrasound probe (Sun and Anthony, 2012; Sun et al., 2013a,b). The system comprises: (1) an optical camera kinematically coupled to an ultrasound transducer, (2) software that extracts unique skin features, and (3) software that localizes the transducer by tracking the skin features (Sun and Anthony, 2012; Sun et al., 2013a,b, 2014).

The process involves simultaneously recording, synchronously in time, ultrasound probe images inside the body and camera images of the skin. Skin features are used to track relative motion along the body. The same skin features can be used to virtually identify and tag unique locations on the body; this system permits placement of the ultrasound transducer in the same body location and orientation by different operators on serial examinations. In practice, multiple sources of information, e.g., camera images and ultrasound images and known regularities motion smoothness, are combined for optimal probe pose estimation. In this way, a body-contoured skin map is constructed and each ultrasound image is accurately localized with respect to the map—a process termed simultaneous localization and mapping (SLAM; Thrun and Leonard, 2008).

This process results in a spatially accurate three-dimensional (3D) ultrasound volume derived from a standard optically tracked two-dimensional sonogram. Consequently, an ultrasound probe may be approximately registered/relocated at a known location and orientation with respect to patient coordinates. The 3D volume could be used to generate multi-planar images which should be nearly indistinguishable from directly acquired ultrasound images (Sun et al., 2014). Sonograms can be reformatted in any plane, allowing comparison of matched images across serial examinations. The re-registration ability should improve the ability to longitudinally monitor near-surface or fixed location organs or tissue structures.

Along with improvement of the hardware technologies and better integration into systems, significant research is needed to further develop, refine, and validate analysis algorithms and workflow software. These improvements would support the effort to produce automated or augmented quantitative analyses of ultrasound imagery applied to skeletal muscles. The rapid technological advancements in ultrasound systems will continue to have a meaningful impact on the assessment of muscle quality, and perhaps expand the methods used for the screening and staging of sarcopenia and other neuromuscular disorders.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest. The reviewer ZL and handling Editor declared their shared affiliation, and the handling Editor states that the process nevertheless met the standards of a fair and objective review.