This study describes how to obtain high quality musculoskeletal images using the extended field-of-view ultrasound (EFOV-US) method for the purpose of making muscle fascicle length measures. We apply this method to muscles with fascicles that extend past the field-of-view of common traditional ultrasound (T-US) probes.
Muscle fascicle length, which is commonly measured in vivo using traditional ultrasound, is an important parameter defining a muscle’s force generating capacity. However, over 90% of all upper limb muscles and 85% of all lower limb muscles have optimal fascicle lengths longer than the field-of-view of common traditional ultrasound (T-US) probes. A newer, less frequently adopted method called extended field-of-view ultrasound (EFOV-US) can enable direct measurement of fascicles longer than the field-of-view of a single T-US image. This method, which automatically fits together a sequence of T-US images from a dynamic scan, has been demonstrated to be valid and reliable for obtaining muscle fascicle lengths in vivo. Despite the numerous skeletal muscles with long fascicles and the validity of the EFOV-US method for making measurements of such fascicles, few published studies have utilized this method. In this study, we demonstrate both how to implement the EFOV-US method to obtain high quality musculoskeletal images and how to quantify fascicle lengths from those images. We expect that this demonstration will encourage the use of the EFOV-US method to increase the pool of muscles, both in healthy and impaired populations, for which we have in vivo muscle fascicle length data.
Fascicle length is an important parameter of skeletal muscle architecture, which overall is indicative of a muscle’s ability to produce force1,2. Specifically, a muscle’s fascicle length provides insight into the absolute range of lengths over which a muscle can generate active force3,4. For example, given two muscles with identical values for all isometric force-generating parameters (i.e., average sarcomere length, pennation angle, physiological cross sectional area, contraction state, etc.) except for fascicle length, the muscle with the longer fascicles would produce its peak force at a longer length and would produce force over a wider range of lengths than the muscle with shorter fascicles3. Quantification of muscle fascicle length is important for understanding both healthy muscle function and changes in a muscle’s force-generating capacity, which can occur as a result of altered muscle use (e.g., immobilization5,6, exercise intervention7,8,9, high heel wearing10) or a change in the muscle’s environment (e.g., tendon transfer surgery11, limb distraction12). Measurements of muscle fascicle length were originally obtained through ex vivo cadaveric experiments that allow for direct measurement of dissected fascicles13,14,15,16. The valuable information provided by these ex vivo experiments led to an interest in implementing in vivo methods17,18,19 to address questions that could not be answered in cadavers; in vivo methods allow for quantification of muscle parameters in a native state as well as at different joint postures, different muscle contraction states, different loading or unloading states, and across populations with differing conditions (i.e. healthy/injured, young/old, etc.). Most frequently, ultrasound is the method employed to obtain in vivo muscle fascicle lengths18,19,20; it is quicker, less expensive, and easier to implement than other imaging techniques, such as diffusion tensor imaging (DTI)18,21.
Extended field-of-view ultrasound (EFOV-US) has been demonstrated to be a valid and reliable method for measuring muscle fascicle length in vivo. While commonly implemented, traditional ultrasound (T-US) has a field-of-view which is limited by the ultrasound transducer’s array length (typically between 4 and 6 cm, although there are probes that extend to 10 cm10)18,20. To overcome this limitation, Weng et al. developed an EFOV-US technology that automatically acquires a composite, two-dimensional “panoramic” image (up to 60 cm long) from a dynamic, extended distance scan22. The image is created by fitting together, in real-time, a sequence of traditional, B-mode ultrasound images as the transducer dynamically scans the object of interest. Because sequential T-US images have large overlapping regions, the small differences from one image to the next can be used to calculate the probe motion without the use of external motion sensors. Once the probe motion between two consecutive images is calculated, the “current” image is merged successively with the preceding images. The EFOV-US method allows direct measurement of long, curved muscle fascicles and has been demonstrated to be reliable across muscles, trials, and sonographers23,24,25 and valid for both flat and curved surfaces23,26.
Implementing ultrasound to measure muscle fascicle length in vivo is not trivial. Unlike other imaging techniques that involve more automated protocols (i.e., MRI, CT), ultrasound is dependent on sonographer skill and anatomical knowledge27,28. There is concern that probe misalignment with the fascicle plane may cause substantial error in fascicle measures. One study demonstrates little difference (on average < 3 mm) in measures of fascicle length taken using ultrasound and DTI MRI but also shows that measurement precision is low (standard deviation of difference ~12 mm)29. Still, it has been shown that a novice sonographer, with practice and guidance from an experienced sonographer, can obtain valid meaures using EFOV-US23. Thus, efforts should be made to demonstrate appropriate protocols to reduce human error and improve accuracy of measurements obtained using EFOV-US. Ultimately, developing and sharing appropriate protocols may expand the number of experimenters and laboratories that can reproduce fascicle length data from the literature or obtain novel data in muscles which have not yet been studied in vivo.
In this protocol, we demonstrate how to implement the EFOV-US method to obtain high quality musculoskeletal images that can be used to quantify muscle fascicle length. Specifically, we address (a) collecting EFOV-US images of a single upper limb and a single lower limb muscle (b) determining, in real-time, the “quality” of the EFOV-US image, and (c) quantifying muscle architecture parameters offline. We provide this detailed guide to encourage the adoption of the EFOV-US method for obtaining muscle fascicle length data in muscles that have gone unstudied in vivo due to their long fascicles.
Northwestern University’s Institutional Review Board (IRB) approved the procedures of this study. All participants enrolled in this work gave informed consent prior to beginning the protocol detailed below.
NOTE: The specific ultrasound system used in this study had EFOV-US capabilities and was adopted because we were able to review details about and validity assessments for the algorithm in the scientific literature22,26; multiple other systems with EFOV-US also exist18,20,30. A linear array transducer 14L5 (frequency bandwidth 5-14 MHz) was utilized. The muscles imaged in this protocol are just a small subset of muscles for which US images have been captured and fascicle lengths measured (e.g., triceps25, extensor carpi ulnaris23, medial gastrocnemius10, vastus lateralis24, biceps femoris8,31). This protocol is intended to provide pointers and describe the necessary standards so that that it may be applied to muscles beyond the two examples we provide.
1. Collecting EFOV-US images of muscles
Preparation
Image Acquisition
2. Determining “quality” of the EFOV-US image
3. Quanitfying Muscle Fascicle Length
Extended field-of-view ultrasound (EFOV-US) was implemented to obtain images from the long head of the biceps brachii and the tibialis anterior in 4 healthy volunteers (Table 1). Figure 1 shows what EFOV-US images of both muscles imaged in this representative imaging session and highlights important aspects of each image such as muscle aponeurosis, central tendon, fascicle path, etc. After the imaging session was over, 3 qualitatively “good” images (Figure 2) were analyzed for each muscle in each individual. ImageJ was implemented to measure 4 fascicles per image. In each image, fascicles with paths that could be convincingly visualized from origin to insertion and that were located in different portions of the muscle selected were measured. The average fascicle lengths obtained in this study for the biceps brachii (14.6 ± 1.7 cm) and the tibialis anterior (7.3 ± 0.6 cm) are within the range of fascicle lengths reported previously25,42 (Table 1).
As most challenging and subjective parts of this protocol is determining factors which lead to correctly deeming an image as qualitatively “good” or qualitatively “bad”. We provide several examples of “good” and “bad” images (Figure 2) and how image landmarks and quality vary across people (Figure 3). In addition we have highlighted the portions of the images which are specifically “bad”.
Subject | Gender | Height (m) | Age | Bicep Side | Bicep Fascicle Length (cm) | Tibialis Anterior Side | Tibialis Anterior Fascicle length (cm) |
1 | M | 1.78 | 24 | L | 16.4 ± 0.3 | L | 7.6 ± 0.1 |
2 | F | 1.8 | 23 | R | 12.2 ± 0.2 | L | 7.5 ± 0.2 |
3 | M | 1.82 | 24 | L | 14.9 ± 0.2 | R | 7.7 ± 0.1 |
4 | F | 1.79 | 28 | R | 14.7 ± 0.2 | L | 6.4 ± 0.3 |
Average | 14.6 | 7.3 | |||||
SD | 1.7 | 0.6 |
Table 1: Participant Demographics and Data. Measurements of fascicle length are represented as average ± standard deviation.
Figure 1: Schematic and EFOV images of two example muscles. (left) Illustration of the muscle being studied. (right) Example of “good” images on top and the same image with whole muscle (dark blue), central tendon (light blue), and muscle fascicles (white) outlined. Each image has a corresponding 1 cm scale bar (white) on the bottom right of the image. Please click here to view a larger version of this figure.
Figure 2: Demonstration of Image Quality. Demonstration of three qualitatively “good” and three qualitatively “bad” images obtained from the biceps brachii and the tibialis anterior of participants 1 and 2. (Top A & B) In all the qualitatively “good” images fascicles which extend from internal tendon to muscle aponeurosis can be visualized. We illustrate images which are qualitatively “bad” and should not be analyzed. Portions of the image which qualify it is as “bad” are emphasized (blue boxes and arrows) and include jagged or broken images, excessive or non-anatomically relevant bending, images which exclude the entire fascicle, and images with blurred central tendons. Each image has a scale bar (white vertical line) which represents 1 cm. This portion of the figure is highlighting the variability among images due mainly to the sonographer’s inconsistency across separate imaging sweeps. (Bottom A & B) One “good” biceps and one “good” tibialis anterior muscle are shown. The orange box on the original image is then blown up to illustrate more accurately the zoom that is seen when measuring fascicles in ImageJ. The bottom image shows representative outlined fascicles (white dashed lines). These images are deemed “good” because fascicles can be followed from origin to insertion and the zoomed portion of the image doesn't have substantial distortions or artifacts. Please click here to view a larger version of this figure.
Figure 3: Variability in image quality across individuals. Variability in image quality and visibility exists between participants, largely due to anatomical variation (i.e. muscle size, muscle length, subcutaneous fat content) and differences in muscle content (i.e. amounts of intramuscular fat, connective tissue, fibrosis). Specifically, variations in muscle content and layers of tissue above the muscle can affect the echo intensity of the imaged muscle43. Natural anatomical differences across individuals will result in muscle architectural features varying in location and/or relative size across US images of different individuals. This demonstration of muscles in different participants stresses the importance of a thorough understanding of anatomy and sufficient practice obtaining images on various individuals for gaining confidence in the quality and accuracy of the images being obtained. Please click here to view a larger version of this figure.
Critical steps in the protocol.
There are a few critical components to obtaining quality EFOV-US images that yield valid and reliable fascicle length measures. First, as indicated in method 1.1.2 it is essential that the sonographer take time to become familiar with the anatomy of the muscle being imaged as well as surrounding muscles, bones, and other soft tissue structures. This will improve the sonographer’s ability to image the correct muscle and determine if multiple images are capturing the same muscle plane. Second, the sonographer should practice the protocol on phantoms and multiple pilot participants before collecting data for publication. Ultrasound is known to result in measurement error if the sonographer does not properly identify the fascicle plane, a task which is challenging and can improve with practice. Last, it is strongly recommended to ensure that the validity of the measurements made by the EFOV-US algorithm in the ultrasound system being used has been established. If the method’s accuracy has not been demonstrated, validation can be done easily using an ultrasound phantom23,26 or through comparison to another imaging tool44 or cadaveric dissection45.
Modifications and troubleshooting of the method.
If image visibility is poor or the probe motion is uneven during dynamic scanning, adding ultrasound gel may enhance image quality by improving transducer-to-skin coupling. If image acquisition is cut-off by the algorithm before the entire object of interest is captured, the depth of the image should be increased. Increasing the depth of the image expands the available scanning distance, thus enabling longer objects to be captured within a single EFOV-US image. In general, it is best to refer to the ultrasound system’s manual when attempting to improve or troubleshoot image quality or image acquisition.
Here, we demonstrate how to capture EFOV-US images of the entire muscle from the muscle tendon junction of the origin tendon to the insertion tendon. Capturing the entire muscle is necessary for some muscles, such as the biceps brachii, whose fascicles span nearly the entire length of the muscle. However, for other muscles, such as the tibialis anterior or other pennated muscles, shorter scans that do not include the full muscle belly may still capture entire muscle fascicles. For novice sonographers, acquiring images from shorter scans that still capture full fascicle lengths may decrease the chances of probe misalignment with the fascicle plane and improve image quality, decreasing the potential for fascicle measurement error.
Limitations of the Method
Notably, muscle activation can change muscle fascicle length. Due to the nature of the scanning method, the major limitation of EFOV-US is that it cannot be implemented to study muscle fascicle changes due to dynamic muscle contraction (e.g., during walking46,47). Additionally, due to the time required to capture an EFOV-US image, imaging a muscle at maximum contraction is likely infeasible due to muscle fatigue. Instead, the EFOV-US method is beneficial for sub-maximal or passive imaging. One way to ensure muscle activity is constant across participants, limbs, or sessions is to simultaneously measure EMG during imaging and analyze only images which are taken when the muscle is at some desired activity level. Though recommended, particularly if studying populations with altered neural drive, measures of EMG were not taken in the population studied here.
Though traditional ultrasound has been shown to be valid and reliable for measuring in vivo muscle fascicle lengths, some fascicle measurement error will occur if the sonographer’s alignment of the ultrasound transducer deviates from the fascicle plane27,29,48. Due to the nature of the EFOV-US’s dynamic scan, there is concern that the EFOV-US method may have more error than T-US21,24. While a recent study demonstrated that fascicle measurement error from probe misalignment was not larger in EFOV-US than in the well-established, T-US method23 in a single wrist muscle, a general limitation of B-mode ultrasound is that you are only able to capture a relatively small, 2-dimensional (2D) view of the muscle. The true path of individual fascicles may be 3D; concerns remain that errors associated with measuring lengths of potentially 3D paths from 2D views may be bigger for longer fascicles.
Significance of the Method with Respect to Existing/Alternative Methods
Static, B-mode ultrasound is a widely accepted method for measuring muscle fascicle lengths in vivo. However, the field-of-view of T-US probes limits the length of fascicles that can be directly measured. Instead, measurement of fascicles longer than the field-of-view of T-US requires trigonometric estimation methods, diffusion tensor imaging (DTI), or EFOV-US20. In general, ultrasound imaging is favored over magnetic resonance imaging (MRI) techniques such as DTI because MRI is more expensive and challenging to implement18. Fascicle lengths captured with EFOV-US have been shown to be more accurate than trigonometric estimation methods24,36, which is expected since muscle fascicles regularly follow a curved path, but trigonometric estimation methods assume linearity in their calculation of muscle fascicle length.
It should be noted that though most ultrasound probes are 4-6 cm in length, ultrasound probes up to 10 cm have been used9,10. The 10 cm probes enable a wider field-of-view, enabling the capture of longer, straight fascicles. Still, the longer probe length decreases frame rate, would require the imaging surface (the body) to also be straight to avoid uneven compression of the imaged tissue, and may not be able to capture longer curved fascicles (without the use of EFOV)20.
Future Applications or Directions of the Method
The guide detailed here for obtaining quality EFOV-US images for measuring muscle fascicle length is intended to encourage the use of the EFOV-US method to expand the pool of muscles for which the field has in vivo muscle architecture data. The expectation is that this method be applied to both healthy and impaired populations (e.g., individuals post-stroke38,49 or post-orthopedic surgery) to better understand muscle function and muscle adaptation. In addition, these in vivo data are important for development of models that more accurately predict human movement as well as the development of subject specific musculoskeletal models.
Notably, the EFOV-US method is not limited to measurements of muscle fascicle length. The method has been used for measurement of tendon length50,51 and muscle anatomical cross-sectional area,52,53 as well as for documentation of various superficial lesions54,55. Thus, there is opportunity to develop guides, similar to the one presented here, for obtaining high quality images with the EFOV-US method for various applications.
The authors have nothing to disclose.
We would like to thank Vikram Darbhe and Patrick Franks for their experimental guidance. This work is supported by the National Science Foundation Graduate Research Fellowship Program under Grant No. DGE-1324585 as well as NIH R01D084009 and F31AR076920. Any opinions, findings, and conclusions or recommendations expressed in this material are those of the authors and do not necessarily reflect the views of the National Science Foundation or NIH.
14L5 linear transducers | Siemens | 10789396 | |
Acuson S2000 Ultrasound System | Siemens | 10032746 | |
Adjustable chair (Biodex System) | Biodex Medical Systems | System Pro 4 | |
Skin Marker Medium Tip | SportSafe | n/a | Multi-color 4 Pack recommended |
Ultrasound Gel – Standard 8 Ounce Non-Sterile Fragrance Free Glacial Tint | MediChoice, Owens &Minor | M500812 |