scholarly journals Machine Learning to Extract Muscle Fascicle Length Changes from Dynamic Ultrasound Images in Real-Time

2021 ◽  
Author(s):  
Luis G. Rosa ◽  
Jonathan S. Zia ◽  
Omer T. Inan ◽  
Gregory S. Sawicki
Keyword(s):  
Machine Learning ◽  
Real Time ◽  
Muscle Length ◽  
Ground Truth ◽  
Length Change ◽  
Fascicle Length ◽  
Automated Tracking ◽  
Direct Training ◽  
Muscle Fascicle ◽  
Length Changes

AbstractBackground and objectiveDynamic muscle fascicle length measurements through B-mode ultrasound have become popular for the non-invasive physiological insights they provide regarding musculoskeletal structure-function. However, current practices typically require time consuming post-processing to track muscle length changes from B-mode images. A real-time measurement tool would not only save processing time but would also help pave the way toward closed-loop applications based on feedback signals driven by in vivo muscle length change patterns. In this paper, we benchmark an approach that combines traditional machine learning (ML) models with B-mode ultrasound recordings to obtain muscle fascicle length changes in real-time. To gauge the utility of this framework for ‘in-the-loop’ applications, we evaluate accuracy of the extracted muscle length change signals against time-series’ derived from a standard, post-hoc automated tracking algorithm.MethodsWe collected B-mode ultrasound data from the soleus muscle of six participants performing five defined ankle motion tasks: (a) seated, constrained ankle plantarflexion, (b) seated, free ankle dorsi/plantarflexion, (c) weight-bearing, calf raises (d) walking, and then a (e) mix. We trained machine learning (ML) models by pairing muscle fascicle lengths obtained from standardized automated tracking software (UltraTrack) with the respective B-mode ultrasound image input to the tracker, frame-by-frame. Then we conducted hyperparameter optimizations for five different ML models using a grid search to find the best performing parameters for a combination of high correlation and low RMSE between ML and UltraTrack processed muscle fascicle length trajectories. Finally, using the global best model/hyperparameter settings, we comprehensively evaluated training-testing outcomes within subject (i.e., train and test on same subject), cross subject (i.e., train on one subject, test on another) and within/direct cross task (i.e., train and test on same subject, but different task).ResultsSupport vector machine (SVM) was the best performing model with an average r = 0.70 ±0.34 and average RMSE = 2.86 ±2.55 mm across all direct training conditions and average r = 0.65 ±0.35 and average RMSE = 3.28 ±2.64 mm when optimized for all cross-participant conditions. Comparisons between ML vs. UltraTrack (i.e., ground truth) tracked muscle fascicle length versus time data indicated that ML tracked images reliably capture the salient qualitative features in ground truth length change data, even when correlation values are on the lower end. Furthermore, in the direct training, calf raises condition, which is most comparable to previous studies validating automated tracking performance during isolated contractions on a dynamometer, our ML approach yielded 0.90 average correlation, in line with other accepted tracking methods in the field.ConclusionsBy combining B-mode ultrasound and classical ML models, we demonstrate it is possible to achieve real-time tracking of human soleus muscle fascicles across a number of functionally relevant contractile conditions. This novel sensing modality paves the way for muscle physiology in-the-loop applications that could be used to modify gait via biofeedback or unlock novel wearable device control techniques that could enable restored or augmented locomotion performance.

scholarly journals Machine learning to extract muscle fascicle length changes from dynamic ultrasound images in real-time

PLoS ONE ◽  
2021 ◽  
Vol 16 (5) ◽  
pp. e0246611
Author(s):  
Luis G. Rosa ◽  
Jonathan S. Zia ◽  
Omer T. Inan ◽  
Gregory S. Sawicki
Keyword(s):  
Machine Learning ◽  
Real Time ◽  
Muscle Length ◽  
Ground Truth ◽  
Length Change ◽  
Fascicle Length ◽  
Automated Tracking ◽  
Direct Training ◽  
Muscle Fascicle ◽  
Length Changes

Background and objective Dynamic muscle fascicle length measurements through B-mode ultrasound have become popular for the non-invasive physiological insights they provide regarding musculoskeletal structure-function. However, current practices typically require time consuming post-processing to track muscle length changes from B-mode images. A real-time measurement tool would not only save processing time but would also help pave the way toward closed-loop applications based on feedback signals driven by in vivo muscle length change patterns. In this paper, we benchmark an approach that combines traditional machine learning (ML) models with B-mode ultrasound recordings to obtain muscle fascicle length changes in real-time. To gauge the utility of this framework for ‘in-the-loop’ applications, we evaluate accuracy of the extracted muscle length change signals against time-series’ derived from a standard, post-hoc automated tracking algorithm. Methods We collected B-mode ultrasound data from the soleus muscle of six participants performing five defined ankle motion tasks: (a) seated, constrained ankle plantarflexion, (b) seated, free ankle dorsi/plantarflexion, (c) weight-bearing, calf raises (d) walking, and then a (e) mix. We trained machine learning (ML) models by pairing muscle fascicle lengths obtained from standardized automated tracking software (UltraTrack) with the respective B-mode ultrasound image input to the tracker, frame-by-frame. Then we conducted hyperparameter optimizations for five different ML models using a grid search to find the best performing parameters for a combination of high correlation and low RMSE between ML and UltraTrack processed muscle fascicle length trajectories. Finally, using the global best model/hyperparameter settings, we comprehensively evaluated training-testing outcomes within subject (i.e., train and test on same subject), cross subject (i.e., train on one subject, test on another) and within/direct cross task (i.e., train and test on same subject, but different task). Results Support vector machine (SVM) was the best performing model with an average r = 0.70 ±0.34 and average RMSE = 2.86 ±2.55 mm across all direct training conditions and average r = 0.65 ±0.35 and average RMSE = 3.28 ±2.64 mm when optimized for all cross-participant conditions. Comparisons between ML vs. UltraTrack (i.e., ground truth) tracked muscle fascicle length versus time data indicated that ML tracked images reliably capture the salient qualitative features in ground truth length change data, even when correlation values are on the lower end. Furthermore, in the direct training, calf raises condition, which is most comparable to previous studies validating automated tracking performance during isolated contractions on a dynamometer, our ML approach yielded 0.90 average correlation, in line with other accepted tracking methods in the field. Conclusions By combining B-mode ultrasound and classical ML models, we demonstrate it is possible to achieve real-time tracking of human soleus muscle fascicles across a number of functionally relevant contractile conditions. This novel sensing modality paves the way for muscle physiology in-the-loop applications that could be used to modify gait via biofeedback or unlock novel wearable device control techniques that could enable restored or augmented locomotion performance.

scholarly journals Muscle spindles in human tibialis anterior encode muscle fascicle length changes

2017 ◽  
Vol 117 (4) ◽  
pp. 1489-1498 ◽  
Author(s):  
James Day ◽  
Leah R. Bent ◽  
Ingvars Birznieks ◽  
Vaughan G. Macefield ◽  
Andrew G. Cresswell
Keyword(s):  
Muscle Spindle ◽  
Tibialis Anterior ◽  
Tibialis Anterior Muscle ◽  
Muscle Length ◽  
Muscle Spindles ◽  
Muscle Spindle Afferents ◽  
Fascicle Length ◽  
Muscle Fascicle ◽  
Length Changes

Muscle spindles provide exquisitely sensitive proprioceptive information regarding joint position and movement. Through passively driven length changes in the muscle-tendon unit (MTU), muscle spindles detect joint rotations because of their in-parallel mechanical linkage to muscle fascicles. In human microneurography studies, muscle fascicles are assumed to follow the MTU and, as such, fascicle length is not measured in such studies. However, under certain mechanical conditions, compliant structures can act to decouple the fascicles, and, therefore, the spindles, from the MTU. Such decoupling may reduce the fidelity by which muscle spindles encode joint position and movement. The aim of the present study was to measure, for the first time, both the changes in firing of single muscle spindle afferents and changes in muscle fascicle length in vivo from the tibialis anterior muscle (TA) during passive rotations about the ankle. Unitary recordings were made from 15 muscle spindle afferents supplying TA via a microelectrode inserted into the common peroneal nerve. Ultrasonography was used to measure the length of an individual fascicle of TA. We saw a strong correlation between fascicle length and firing rate during passive ankle rotations of varying rates (0.1–0.5 Hz) and amplitudes (1–9°). In particular, we saw responses observed at relatively small changes in muscle length that highlight the sensitivity of the TA muscle to small length changes. This study is the first to measure spindle firing and fascicle dynamics in vivo and provides an experimental basis for further understanding the link between fascicle length, MTU length, and spindle firing patterns. NEW & NOTEWORTHY Muscle spindles are exquisitely sensitive to changes in muscle length, but recordings from human muscle spindle afferents are usually correlated with joint angle rather than muscle fascicle length. In this study, we monitored both muscle fascicle length and spindle firing from the human tibialis anterior muscle in vivo. Our findings are the first to measure these signals in vivo and provide an experimental basis for exploring this link further.

scholarly journals Functional properties of human muscle spindles

2018 ◽  
Vol 120 (2) ◽  
pp. 452-467 ◽  
Author(s):  
Vaughan G. Macefield ◽  
Thomas P. Knellwolf
Keyword(s):  
Functional Properties ◽  
Somatosensory System ◽  
Muscle Length ◽  
Muscle Spindles ◽  
Human Muscle ◽  
Independent Control ◽  
Fascicle Length ◽  
Muscle Fascicle ◽  
The Central Nervous System ◽  
Human Muscles

Muscle spindles are ubiquitous encapsulated mechanoreceptors found in most mammalian muscles. There are two types of endings, primary and secondary, and both are sensitive to changes in muscle length and velocity, with the primary endings having a greater dynamic sensitivity. Unlike other mechanoreceptors in the somatosensory system, muscle spindles are unique in possessing motor innervation, via γ-motoneurons (fusimotor neurons), that control their sensitivity to stretch. Much of what we know about human muscles spindles comes from studying the behavior of their afferents via intraneural microelectrodes (microneurography) inserted into accessible peripheral nerves. We review the functional properties of human muscle spindles, comparing and contrasting with what we know about the functions of muscle spindles studied in experimental animals. As in the cat, many human muscle spindles possess a background discharge that is related to the degree of muscle stretch, but mean firing rates are much lower (~10 Hz). They can faithfully encode changes in muscle fascicle length in passive conditions, but higher level extraction of information is required by the central nervous system to measure changes in muscle length during muscle contraction. Moreover, although there is some evidence supporting independent control of human muscle spindles via fusimotor neurons, any effects are modest compared with the clearly independent control of fusimotor neurons observed in the cat.

scholarly journals Information from dynamic length changes improves reliability of static ultrasound fascicle length measurements

PeerJ ◽  
2017 ◽  
Vol 5 ◽  
pp. e4164 ◽  
Author(s):  
Jeroen Aeles ◽  
Glen A. Lichtwark ◽  
Sietske Lenchant ◽  
Liesbeth Vanlommel ◽  
Tijs Delabastita ◽  
...  
Keyword(s):  
Length Change ◽  
Medial Gastrocnemius ◽  
Ultrasound Images ◽  
Initial Image ◽  
Fascicle Length ◽  
Mean Squared Errors ◽  
Length Measurements ◽  
Length Changes ◽  
Dynamic Length ◽  
Effectiveness Studies

PurposeVarious strategies for improving reliability of fascicle identification on ultrasound images are used in practice, yet these strategies are untested for effectiveness. Studies suggest that the largest part of differences between fascicle lengths on one image are attributed to the error on the initial image. In this study, we compared reliability results between different strategies.MethodsStatic single-image recordings and image sequence recordings during passive ankle rotations of the medial gastrocnemius were collected. Images were tracked by three different raters. We compared results from uninformed fascicle identification (UFI) and results with information from dynamic length changes, or data-informed tracking (DIT). A second test compared tracking of image sequences of either fascicle shortening (initial-long condition) or fascicle lengthening (initial-short condition).ResultsIntra-class correlations (ICC) were higher for the DIT compared to the UFI, yet yielded similar standard error of measurement (SEM) values. Between the initial-long and initial-short conditions, similar ICC values, coefficients of multiple determination, mean squared errors, offset-corrected mean squared errors and fascicle length change values were found for the DIT, yet with higher SEM values and greater absolute fascicle length differences between raters on the first image in the initial-long condition and on the final image in the initial-short condition.ConclusionsDIT improves reliability of fascicle length measurements, without lower SEM values. Fascicle length on the initial image has no effect on subsequent tracking results. Fascicles on ultrasound images should be identified by a single rater and care should be taken when comparing absolute fascicle lengths between studies.

scholarly journals Magnetomicrometry

2021 ◽  
Vol 6 (57) ◽  
pp. eabg0656
Author(s):  
C. R. Taylor ◽  
S. S. Srinivasan ◽  
S. H. Yeon ◽  
M. K. O’Donnell ◽  
T. J. Roberts ◽  
...  
Keyword(s):  
Real Time ◽  
Magnetic Beads ◽  
Control Strategies ◽  
Muscle Length ◽  
Human Movement ◽  
Biological Tissues ◽  
Wearable Robots ◽  
Length Changes ◽  
Portable Sensor

We live in an era of wearable sensing, where our movement through the world can be continuously monitored by devices. Yet, we lack a portable sensor that can continuously monitor muscle, tendon, and bone motion, allowing us to monitor performance, deliver targeted rehabilitation, and provide intuitive, reflexive control over prostheses and exoskeletons. Here, we introduce a sensing modality, magnetomicrometry, that uses the relative positions of implanted magnetic beads to enable wireless tracking of tissue length changes. We demonstrate real-time muscle length tracking in an in vivo turkey model via chronically implanted magnetic beads while investigating accuracy, biocompatibility, and long-term implant stability. We anticipate that this tool will lay the groundwork for volitional control over wearable robots via real-time tracking of muscle lengths and speeds. Further, to inform future biomimetic control strategies, magnetomicrometry may also be used in the in vivo tracking of biological tissues to elucidate biomechanical principles of animal and human movement.

Study on the Length Changes of Bearings in Continuous Motion of Complex Mechanical System Based on ADMAS Simulation

2014 ◽  
Vol 539 ◽  
pp. 29-33
Author(s):  
Ning Li ◽  
Yue Juan Li
Keyword(s):  
Muscle Length ◽  
Surface Force ◽  
Length Change ◽  
Basketball Player ◽  
Validity And Reliability ◽  
Basketball Players ◽  
Concentrated Force ◽  
Virtual Displacement ◽  
Training Time ◽  
Length Changes

In this paper, based on the volume force, surface force, concentrated force of elastic-plastic body, combining with the principle of virtual displacement, we use virtual prototype analysis software ADAMS to design the displacement simulation system of basketball players muscle length change in continued training. In order to verify the validity and reliability of the system, we establish muscle simulation model of basketball player continued training, and input parameters of simulation training time and training task in the model, import the simulation results into the MATLAB software, obtain the curve changes of muscle displacement with time, and combining with the curve get the muscle length changing parameter. It provides a technical reference for the training of basketball players.

Dynamic relations among length, tension, and intracellular Ca2+ in activated ferret papillary muscles

1998 ◽  
Vol 275 (6) ◽  
pp. H1957-H1962 ◽  
Author(s):  
Yasutake Saeki ◽  
Satoshi Kurihara ◽  
Kimiaki Komukai ◽  
Tetsuya Ishikawa ◽  
Kiyohiro Takigiku
Keyword(s):  
Cardiac Troponin ◽  
Muscle Length ◽  
Length Change ◽  
Troponin C ◽  
Cardiac Troponin C ◽  
Mechanical Constraints ◽  
Tension Response ◽  
Cardiac Muscles ◽  
Length Changes ◽  
Cross Bridges

To study the effects of mechanical constraints on the Ca2+ affinity of cardiac troponin C, we analyzed the tension and aequorin light (AL) responses to sinusoidal length changes (5–10% of the initial muscle length) in aequorin-injected, tetanized cardiac muscles. The amplitude of the quasi-sinusoidal tension and AL responses decreased with increasing length-perturbation frequency from 0.5 to 1 Hz at 24°C and from 1 to 3 Hz at 30°C. The increase in AL corresponded well to the decrease in tension; likewise, the decrease in AL to the increase in tension and the tension response lagged behind the length change. A further increase in frequency (>1 Hz at 24°C and >3 Hz at 30°C) markedly increased the amplitude of the tension responses but decreased the amplitude of the AL responses. The increase in AL lagged behind the decrease in tension; likewise, the decrease in AL lagged behind the increase in tension, and the tension response led the length change. From previous mechanistic interpretations of the frequency dependence of the amplitude of tension response, we argue that the Ca2+affinity of cardiac troponin C changes in parallel with the active tension (i.e., the number of active cross bridges) but not with the passive tension produced by the length perturbation-induced cross-bridge strain.

Tibialis anterior muscle fascicle dynamics adequately represent postural sway during standing balance

2013 ◽  
Vol 115 (12) ◽  
pp. 1742-1750 ◽  
Author(s):  
James T. Day ◽  
Glen A. Lichtwark ◽  
Andrew G. Cresswell
Keyword(s):  
Tibialis Anterior ◽  
Tibialis Anterior Muscle ◽  
Sensory Information ◽  
Quiet Standing ◽  
Emg Activity ◽  
Proprioceptive Information ◽  
Fascicle Length ◽  
Muscle Fascicle ◽  
Length Changes

To maintain a stable, upright posture, the central nervous system (CNS) must integrate sensory information from multiple sources and subsequently generate corrective torque about the ankle joint. Although proprioceptive information from the muscles that cross this joint has been shown to be vital in this process, the specific source of this information remains questionable. Recent research has been focused on the potential role of tibialis anterior (TA) muscle during standing, largely due to the lack of modulation of its activity throughout the sway cycle. Ten young, healthy subjects were asked to stand normally under varying conditions, for periods of 60 s. During these trials, intramuscular electromyographic (EMG) activity and the fascicle length of three distinct anatomical regions of TA were sampled synchronously with kinematic data regarding sway position. In the quiet standing conditions, TA muscle activity was unmodulated and fascicle length changes in each region were tightly coupled with changes in sway position. In the active sway condition, more EMG activity was observed in TA and the fascicle length changes were decoupled from sway position. No regional specific differences in correlation values were observed, contrasting previous observations. The ability of the fascicles to follow sway position builds upon the suggestion that TA is well placed to provide accurate, straightforward sensory information to the CNS. As previously suggested, through reciprocal inhibition, afferent information from TA could help to regulate plantar flexor torque at relevant phases of the sway cycle. The proprioceptive role of TA appears to become complicated during more challenging conditions.

scholarly journals The series-elastic shock absorber: tendons attenuate muscle power during eccentric actions

2010 ◽  
Vol 109 (2) ◽  
pp. 396-404 ◽  
Author(s):  
Thomas J. Roberts ◽  
Emanuel Azizi
Keyword(s):  
Muscle Power ◽  
Power Input ◽  
Fascicle Length ◽  
Lever System ◽  
Muscle Fascicle ◽  
Energy Conserving ◽  
Length Changes ◽  
Muscle Fascicles ◽  
Energy Absorbing ◽  
Elastic Elements

Elastic tendons can act as muscle power amplifiers or energy-conserving springs during locomotion. We used an in situ muscle-tendon preparation to examine the mechanical function of tendons during lengthening contractions, when muscles absorb energy. Force, length, and power were measured in the lateral gastrocnemius muscle of wild turkeys. Sonomicrometry was used to measure muscle fascicle length independently from muscle-tendon unit (MTU) length, as measured by a muscle lever system (servomotor). A series of ramp stretches of varying velocities was applied to the MTU in fully activated muscles. Fascicle length changes were decoupled from length changes imposed on the MTU by the servomotor. Under most conditions, muscle fascicles shortened on average, while the MTU lengthened. Energy input to the MTU during the fastest lengthenings was −54.4 J/kg, while estimated work input to the muscle fascicles during this period was only −11.24 J/kg. This discrepancy indicates that energy was first absorbed by elastic elements, then released to do work on muscle fascicles after the lengthening phase of the contraction. The temporary storage of energy by elastic elements also resulted in a significant attenuation of power input to the muscle fascicles. At the fastest lengthening rates, peak instantaneous power input to the MTU reached −2,143.9 W/kg, while peak power input to the fascicles was only −557.6 W/kg. These results demonstrate that tendons may act as mechanical buffers by limiting peak muscle forces, lengthening rates, and power inputs during energy-absorbing contractions.

Sign in / Sign up

Export Citation Format

Share Document