Gait analysis together with musculoskeletal modeling can be used to assess pathological gait, 1 predict musculoskeletal loading, 2 and evaluate the outcome of clinical interventions. 3 The model used for musculoskeletal analyses can be created directly from medical images 4 or
Hans Kainz, Hoa X. Hoang, Chris Stockton, Roslyn R. Boyd, David G. Lloyd, and Christopher P. Carty
Yumeng Li, He Wang, and Kathy J. Simpson
. 16 Assessing tibiofemoral contact forces has been suggested as an essential approach to understand the initiation and progression of knee injuries and diseases. 18 Computer-simulated musculoskeletal models are often used to estimate tibiofemoral contact forces during various movements. 19 , 20
R. Tyler Richardson, Elizabeth A. Rapp, R. Garry Quinton, Kristen F. Nicholson, Brian A. Knarr, Stephanie A. Russo, Jill S. Higginson, and James G. Richards
Musculoskeletal modeling is capable of estimating physiological parameters that cannot be directly measured, 1 , 2 however, the validity of the results must be assessed. A substantial challenge of modeling the shoulder lies in proper implementation of scapular kinematics. 3 , 4 Scapular
David C. Kingston and Stacey M. Acker
musculoskeletal model. Black spheres are manually selected landmarks matching those from Horsman et al. Blue lines indicate muscle paths. Green spheres within a muscle path are scaled VIA points. Red lines are knee joint ligaments (not used in this iteration). The reader is referred to the online version of this
Zachary F. Lerner, Derek J. Haight, Matthew S. DeMers, Wayne J. Board, and Raymond C. Browning
Net muscle moments (NMMs) have been used as proxy measures of joint loading, but musculoskeletal models can estimate contact forces within joints. The purpose of this study was to use a musculoskeletal model to estimate tibiofemoral forces and to examine the relationship between NMMs and tibiofemoral forces across walking speeds. We collected kinematic, kinetic, and electromyographic data as ten adult participants walked on a dual-belt force-measuring treadmill at 0.75, 1.25, and 1.50 m/s. We scaled a musculoskeletal model to each participant and used OpenSim to calculate the NMMs and muscle forces through inverse dynamics and weighted static optimization, respectively. We determined tibiofemoral forces from the vector sum of intersegmental and muscle forces crossing the knee. Estimated tibiofemoral forces increased with walking speed. Peak earlystance compressive tibiofemoral forces increased 52% as walking speed increased from 0.75 to 1.50 m/s, whereas peak knee extension NMMs increased by 168%. During late stance, peak compressive tibiofemoral forces increased by 18% as speed increased. Although compressive loads at the knee did not increase in direct proportion to NMMs, faster walking resulted in greater compressive forces during weight acceptance and increased compressive and anterior/posterior tibiofemoral loading rates in addition to a greater abduction NMM.
Fabien Dal Maso, Mickaël Begon, and Maxime Raison
One approach to increasing the confidence of muscle force estimation via musculoskeletal models is to minimize the root mean square error (RMSE) between joint torques estimated from electromyographic-driven musculoskeletal models and those computed using inverse dynamics. We propose a method that reduces RMSE by selecting subsets of combinations of maximal voluntary isometric contraction (MVIC) trials that minimize RMSE. Twelve participants performed 3 elbow MVIC in flexion and in extension. An upper-limb electromyographic-driven musculoskeletal model was created to optimize maximum muscle stress and estimate the maximal isometric force of the biceps brachii, brachialis, brachioradialis, and triceps brachii. Maximal isometric forces were computed from all possible combinations of flexion-extension trials. The combinations producing the smallest RMSE significantly reduced the normalized RMSE to 7.4% compared with the combination containing all trials (9.0%). Maximal isometric forces ranged between 114–806 N, 64–409 N, 236–1511 N, and 556–3434 N for the brachii, brachialis, brachioradialis, and triceps brachii, respectively. These large variations suggest that customization is required to reduce the difference between models and actual participants’ maximal isometric force. While the smallest previously reported RMSE was 10.3%, the proposed method reduced the RMSE to 7.4%, which may increase the confidence of muscle force estimation.
Roy Müller, Tobias Siebert, and Reinhard Blickhan
In locomotion, humans have to deal with irregularities in the ground. When they encounter uneven terrain with changes in vertical height, they adjust the geometry of their legs. Recent investigations have shown that the preactivation of the gastrocnemius muscle (GM) correlates with the ankle angle at touchdown, but it is as of yet unclear why these adjustments were achieved by the GM and not by the preactivation of the tibialis anterior (TA). To examine the differences between TA regulation and GM regulation regarding (1) ankle angle adjustment and (2) joint stiffness, we used a three-segment musculoskeletal model with two antagonistic muscles (GM, TA). During the GM regulation, the ankle angle was adjusted from 121° to 109° (dorsiflexion) by a 41% decrease in the GM activation. During the TA regulation, the activation of TA must be increased by about 52%. In addition, we found that the ankle stiffness was most sensitive to changes in activation of the GM and decreased by about 20% while adjusting the angle. In contrast, the ankle stiffness remains similar when using TA regulation. Thus, the GM regulation is more adequate for adjustment in the ankle joint, enabling sufficient regulation of angle and stiffness.
Angelica E. Lang, Soo Y. Kim, Stephan Milosavljevic, and Clark R. Dickerson
mechanisms causing kinematic alterations and to identify contributing muscles that may benefit from focused rehabilitation. 9 , 10 Musculoskeletal modeling allows for estimation of muscle forces and loading strategies for more muscles than can be feasibly measured. For a pathological population, the goal of
David Hawkins and Mark Smeulders
The purpose of this study was to determine if the Hill model, used to describe the force-velocity relationship for isolated tetanically stimulated muscle, could be modified and used to describe the torque-velocity behavior of the knee for maximally and submaximally stimulated quadriceps and hamstrings muscles. Fourteen subjects performed both knee flexion and extension movements at 100%, 70%, and 40% of maximum isometric effort. For each effort level, the knee was allowed to move against resistances equal to 75%, 50%, 25%, and 0% of the specified effort level. An electrogoniometer quantified knee angle. Knee velocity was determined by numerically differentiating the joint angle data. Torque-velocity-activation (or effort level) data were determined for each trial. Model parameters were determined to give the best fit to the data for each subject. Average parameter values were determined for each gender and for the entire group. The modified Hill-type model accurately described the relationship between torque, velocity, and muscle activation level for subject-specific parameters but not for parameters averaged across genders or the entire group.
Maarten F. Bobbert, Han Houdijk, Jos J. de Koning, and Gert de Groot
To gain a better understanding of push-off mechanics in speed skating, forward simulations were performed with a model comprising four body segments and six muscles. We started with a simulated maximum height one-legged jump, obtained by optimization of muscle stimulation time histories. The simulated jump was very similar to one-legged jumps produced by a human, indicating that the model was realistic. We subsequently studied how performance was affected by introducing four conditions characteristic of speed skating: (a) We changed the initial position from that in jumping to that at the start of the push-off phase in skating. This change was accommodated by a delay in stimulation onset of the plantar flexors in the optimal solution. (b) The friction between foot and ground was reduced to zero. As a result, maximum jump height decreased by 1.2 cm and performance became more sensitive to errors in muscle stimulation. The reason is that without surface friction, the foot had to be prevented from slipping away, which constrained the solution space and reduced the tolerance to errors in stimulation. (c) We introduced the requirement to maintain the upper body in a more or less horizontal position. This change could be accommodated by a delay in stimulation onset of the hamstrings, which inevitably caused a reduction in maximum jump height by 11.6 cm. (d) We increased the effective foot length from 16.5 cm, representative of jumping, to 20.5 cm, representative of skating with klapskates. At the 20.5-cm foot length, rotation of the foot did not start during the buildup of plantar flexion moment as it did at smaller foot lengths, but was delayed until hip and knee extension moments decreased. This caused an unbalanced increase in segment angular velocities and muscle shortening velocities, leading to a decrease in muscle force and muscle work and a further decrease in maximum jump height by approximately 5 cm. Qualitatively, these findings help clarify why and how performance of speed skaters depends on the location of the hinge of their skate.