Since tendons are passive viscoelastic structures, when they undergo a loading-unloading cycle they must exhibit a negative work loop i. However, prior studies using this Indirect approach report large positive work loops, often estimating that tendons return 2—5 J of elastic energy for every 1 J of energy stored. More direct ultrasound estimates of tendon kinematics have emerged that quantify tendon elongations by tracking either the muscle-tendon junction or localized tendon tissue. However, it is unclear if these yield more plausible estimates of tendon dynamics.
Our objective was to compute tendon work loops and hysteresis losses using these two Direct tendon kinematics estimates during human walking. We interpret this finding to suggest that Direct approaches provide more plausible estimates than the Indirect approach, and may be preferable for understanding tendon energy storage and return. These trends suggest that Direct estimates also contain some level of error, albeit much smaller than Indirect estimates. Overall, this study serves to highlight the complexity and difficulty of estimating tendon dynamics non-invasively, and the care that must be taken to interpret biological function from current ultrasound-based estimates.
This is an open access article distributed under the terms of the Creative Commons Attribution License , which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. Data Availability: All relevant data are within the paper and its Supporting Information files.
The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. There was no additional external funding received for this study. Competing interests: The authors have declared that no competing interests exist. Tendinous tissues throughout the musculoskeletal system perform a variety of important functions to enable economical, safe, and powerful movements. These tissues can serve as an energy-saving mechanism by storing and returning elastic potential energy [ 1 — 6 ], as a safety mechanism for muscles by absorbing energy during impact or landing [ 7 , 8 ], and in certain animals as a power amplification mechanism by overcoming intrinsic limitations on muscle contractile dynamics [ 9 — 11 ].
Tendon mechanical properties e. However, the aforementioned functional roles of tendons must be studied and understood within the context of movement, highlighting the importance of in vivo measurement techniques able to capture tendon dynamics. Ultrasound imaging is increasingly used to quantify tendon dynamics during human movement [ 12 , 13 ]. To quantify muscle and tendon dynamics in non-human experiments, sonomicrometer crystals are often implanted to facilitate precise tracking of tissue length changes; such implants are generally not practical in human experiments.
Instead, human muscle and tendon kinematics are tracked using anatomical feature tracking and a growing number of ultrasound image processing algorithms, for example, based on speckle-tracking [ 14 ] or affine optical flow [ 15 ]. One of the most common ways to estimate tendon kinematics, which we term Indirect , uses ultrasound-based measurements of muscle fascicle length change and pennation angle in conjunction with joint kinematics derived from motion capture Fig 1. Longitudinal length changes of the muscle are subtracted from the overall muscle-tendon unit MTU length to estimate changes in the associated tendon length [ 3 , 17 — 21 ].
Here, MTU length is estimated from previously published regression equations based on joint kinematics; equations which themselves were derived from cadaveric studies [ 22 , 23 ]. This method provides a lumped estimate of tendinous tissue elongation accumulated from both the proximal and distal free tendons and aponeuroses. With this method the ultrasound transducer is positioned over the muscle belly. Right: Indirect tendon kinematics can then be combined with tendon force estimates to examine tendon work loops.
Walking data shown here are from [ 19 ], and calf-raise data are from [ 24 ] see Methods for details on how data were digitized.
ignamant.cl/wp-includes/55/4511-espiar-conversaciones-de.php Tendinous tissue length change can be combined with tendon force estimates to understand the dynamic function of tendons during movement. In prior studies, tendon force estimates have been obtained from percutaneous force transducers [ 19 ], but are more frequently obtained by dividing joint moments estimated via inverse dynamics by the tendon moment arm about the joint [ 17 , 24 ]. Tendon dynamics e. If net negative mechanical work is performed over a loading-unloading cycle, then this is referred to as a negative work loop, and vice versa for net positive work.
For passive structures such as tendons, energy loss due to hysteresis—mechanical energy dissipated as heat during phases of material deformation and recovery—can then be estimated as the area between the tendon loading and unloading curves i. Validating estimates of tendon dynamics is challenging, particularly when combining tendon length change and force estimates to evaluate mechanical work and energy.
This is particularly challenging for in vivo human studies, which generally rely on non-invasive 2D images to estimate length changes of 3D tissues, and which lack comprehensive measurement of individual muscle and tendon forces. One way to assess the validity of in vivo dynamics estimates is to evaluate tendon work loops. Tendinous tissues are passive i. Thus, a simple litmus test i. Regardless of the precise magnitude of hysteresis loss in this range, we would not expect for tendons to exhibit positive work loops.
After all, net positive work produced by a passive tissue would result in a negative hysteresis loss, which does not make physical sense. Net positive work would require an active, power-generating contractile element. In other words, tendon work loops derived from ultrasound imaging should at least confirm that passive, spring-like tendons are indeed behaving in passive, spring-like fashion.
While not intended as a comprehensive validation, this litmus test is an important scientific sanity check, which assists us in gauging our confidence in tendon dynamics derived from ultrasound-based estimates of tendon kinematics.
There is empirical evidence that Indirect estimates of tendon kinematics, when combined with estimates of tendon kinetics, often result in large positive work loops [ 17 , 19 , 24 , 29 , 30 ], which are not physiologically plausible Fig 1 [ 24 ]. For example, tendinous tissues have been estimated to perform 2—3 times as much positive work as negative work over the stance phase of walking i. Taken at face value, this would suggest that passive tendons are acting like motors or muscles , performing net positive work.
This suggests a fundamental problem with Indirect tendon estimates, either due to measurement inaccuracy or incorrect methodological assumptions. Given the growing use of ultrasound imaging to understand the functional role of tendons during dynamic human movement and to translate this understanding, for example into the development of assistive and rehabilitative technologies, it is essential to investigate and resolve this issue.
The purpose of the present report is to provide insight on the current state of the research on tendinopathy and to identify promising areas for future investigation. This article has been cited by other articles in PMC. Articles from BioMed Research International are provided here courtesy of.
Alternative ultrasound methods have emerged for more directly estimating tendon kinematics. A third method, which we term Direct Tendon , uses various speckle-tracking algorithms to quantify local elongations within the tendon [ 14 , 31 — 34 ]. However, it is currently unclear if these Direct estimates also result in physiologically implausible positive work loops, or if these yield tendon results more consistent with expected negative work loops. The purpose of this study was to compute tendon work loops and apparent hysteresis loss using each of these Direct ultrasound estimates of tendon kinematics, and then to compare the results to previously reported values based on Indirect kinematic estimates.
Although methods differ in the portion of tendinous tissue characterized, presumably with differences in absolute length change estimates, we hypothesized that tendon kinematics estimated using each Direct experimental method would yield negative work loops for the Achilles tendon during human walking across a range of speeds. We would interpret this result to be more consistent with expectations for passive tendinous tissues, thereby affirming confidence in using these estimates to understand human movement dynamics.
The experimental protocol was approved by, and all subjects provided written informed consent according to, the University of Wisconsin Health Sciences Internal Review Board. Accordingly, our cohort was not based on an a priori power analysis specific to the hypotheses presented herein. To briefly summarize, subjects walked barefoot on a dual-belt, force-sensing treadmill for 2 minutes at each of three walking speeds 0. Subjects then completed two 2-minute walking trials at each of these three walking speeds in a randomized block design, one trial for each of the two ultrasound imaging locations described below.
Complete measurement and analysis details are provided in Franz et al. We collected human motion and force data using standard gait analysis procedures, and employed previously-published ultrasound measurement techniques to record 2D tendon kinematics.
During each trial, we collected 3D pelvis and lower-limb kinematics at Hz using an optical motion capture system Motion Analysis, Corp. The sampling frequency for these measurements was depth-dependent. We collected the position and orientation of the ultrasound transducer using three retroreflective markers placed on the custom orthotic. For each of these two imaging locations, we recorded 5 strides per condition distributed over each 2 minute walking trial. We later identified these strides in our motion capture data. Using an analog sync signal emitted from the Ultrasonix system, the ultrasound and motion data were synchronized to within 5 ms i.
We analyzed each series of ultrasound data in two phases. Phase I consisted of estimating local tissue displacements from the ultrasound recordings i. The Direct Tendon method used a 2D ultrasound speckle-tracking algorithm to track longitudinal free Achilles tendon tissue displacements based on previously published and validated techniques [ 36 , 37 ]. The region of interest width ensured that only tendinous tissue was included in the tracking routine. Peaks of these frame-to-frame cross-correlations defined the gross nodal displacements which were refined by estimating subpixel displacements using the peaks of a 2D quartic spline surface fit to the correlation functions.
Final nodal trajectories were determined as the weighted average of forward and backward tracking results assuming cyclic motion trajectories. For the present analysis, we used the average of all nodal trajectories as a gross estimate of Achilles tendon tissue displacement over a walking stride. In this method the ultrasound transducer is positioned over the MTJ. In this method the ultrasound transducer is positioned over the Achilles free tendon, distal to the soleus MTJ. Right: This Direct Tendon method yields negative tendon work loops on average, for all walking speeds tested.
We calculated net ankle moment via standard inverse dynamics analysis and used this to approximate Achilles tendon force, as detailed below. Marker trajectories and ground reaction force measurements were fourth-order low-pass filtered at 6 Hz and Hz cutoff frequencies, respectively. We assumed that the net ankle extensor moment was borne completely by forces generated by the triceps surae soleus, medial and lateral gastrocnemius muscles and transmitted through the Achilles free tendon.
Accordingly, we calculated the instantaneous Achilles tendon force by dividing the net ankle moment by subject-specific measures of the Achilles tendon moment arm. Combining estimates of tendon kinematics from ultrasound and kinetics from inverse dynamics , we then calculated stance phase tendon work loops tendon force versus tendon length change , derived from both Direct MTJ and Direct Tendon approaches.
Next, we integrated these tendon work loops to calculate stance phase positive work during tendon loading, in Joules and negative work during unloading. Net work was then computed by summing negative work energy absorbed during tendon loading and positive work energy returned during tendon unloading.
Net work provides an estimate of the hysteresis loss of the tendon. However, to facilitate comparisons between walking speeds we divided this value by the negative work tendon energy stored and report hysteresis loss as a percentage, a common convention in biomechanics e. Net tendon work and hysteresis loss from Direct MTJ vs. Hysteresis loss was estimated from previously published studies that computed tendon kinematics via the Indirect method, and which also reported an estimate of tendon force.
Several previous studies have presented figures showing experimentally-estimated Achilles tendon work loops e. In both cases these are sufficient data to compute tendon hysteresis loss; however, hysteresis losses were not explicitly computed nor reported in these articles. Data digitization involves: i defining the directions of x and y axes on the figure, ii calibrating these axes using numerical axis labels in the publication to convert between physical distances on the page and actual units of tendon force or length, iii manually tracing each data curve by selecting discrete points along it to capture a series of x and y data points, then iv exporting these data points in units of force or length based on the established calibration for post-processing.
From Sakuma et al. These data were imported into Matlab, where we integrated the work loops to calculate positive and negative work, then used these work values to compute tendon hysteresis loss identical to calculations performed for Direct MTJ and Direct Tendon methods.