Prediction of ground reaction forces during gait based on kinematics and a neural network model

Kinetic information during human gait can be estimated with inverse dynamics, which is based on anthropometric, kinematic, and ground reaction data. While collecting ground reaction data with a force plate is useful, it is costly and requires regulated space. The goal of this study was to propose a new, accurate methodology for predicting ground reaction forces (GRFs) during level walking without the help of a force plate. To predict GRFs without a force plate, the traditional method of Newtonian mechanics was used for the single support phase. In addition, an artificial neural network (ANN) model was applied for the double support phase to solve statically indeterminate structure problems. The input variables of the ANN model, which were selected to have both dependency and independency, were limited to the trajectory, velocity, and acceleration of the whole segment's mass centre to minimise errors. The predicted GRFs were validated with actual GRFs through a ten-fold cross-validation method, and the correlation coefficients (R) for the ground forces were 0.918 in the medial–lateral axis, 0.985 in the anterior–posterior axis, and 0.991 in the vertical axis during gait. The ground moments were 0.987 in the sagittal plane, 0.841 in the frontal plane, and 0.868 in the transverse plane during gait. The high correlation coefficients(R) are due to the improvement of the prediction rate in the double support phase. This study also proved the possibility of calculating joint forces and moments based on the GRFs predicted with the proposed new hybrid method. Data generated with the proposed method may thus be used instead of raw GRF data in gait analysis and in calculating joint dynamic data using inverse dynamics.     

Ground reaction forces during human locomotion on railroad ballast

Locomotion over ballast surfaces provides a unique situation for investigating the biomechanics of gait. Although much research has focused on level and sloped walking on a smooth, firm surface in order to understand the common kinematic and kinetic variables associated with human locomotion, the literature currently provides few if any discussions regarding the dynamics of locomotion on surfaces that are either rocky or uneven. The purpose of this study was to investigate a method for using force plates to measure the ground reaction forces (GRFs) during gait on ballast. Ballast is a construction aggregate of unsymmetrical rock used in industry for the purpose of forming track bed on which railway ties are laid or in yards where railroad cars are stored. It is used to facilitate the drainage of water and to create even running surfaces. To construct the experimental ballast surfaces, 31.75 mm (1 1/4 in.) marble ballast at depths of approximately 63.5 mm (2.5 in.) or 101.6 mm (4 in.) were spread over a carpeted vinyl tile walkway specially designed for gait studies. GRF magnitudes and time histories from a force plate were collected under normal smooth surface and under both ballast surface conditions for five subjects. GRF magnitudes and time histories during smooth surface walking were similar to GRF magnitudes and time histories from the two ballast surface conditions. The data presented here demonstrate the feasibility of using a force plate system to expand the scope of biomechanical analyses of locomotion on ballast surfaces.     

Estimation of unmeasured ground reaction force data based on the oscillatory characteristics of the center of mass during human walking

To enhance the wearability of portable motion-monitoring devices, the size and number of sensors are minimized, but at the expense of quality and quantity of data collected. For example, owing to the size and weight of low-frequency force transducers, most currently available wearable gait measurement systems provide only limited, if any, elements of ground reaction force (GRF) data. To obtain the most GRF information possible with a minimal use of sensors, we propose a GRF estimation method based on biomechanical knowledge of human walking. This includes the dynamics of the center of mass (CoM) during steady human gait resembling the oscillatory behaviors of a mass-spring system. Available measurement data were incorporated into a spring-loaded inverted pendulum with translating pivot. The spring stiffness and simulation parameters were tuned to match, as accurately as possible, the available data and oscillatory characteristics of walking. Our results showed that the model simulation estimated reasonably well the unmeasured GRF. Using the vertical GRF and CoP profile for gait speeds ranging from 0.93 to 1.89 m/s, the anterior-posterior (A-P) GRF was estimated and resulted in an average correlation coefficient of R = 0.982 ± 0.009. Even when the ground contact timing and gait speed information were alone available, our method estimated GRFs resulting in R = 0.969 ± 0.022 for the A-P and R = 0.891 ± 0.101 for the vertical GRFs. This research demonstrates that the biomechanical knowledge of human walking, such as inherited oscillatory characteristics of the CoM, can be used to gain unmeasured information regarding human gait dynamics.     

A mobile gait monitoring system for abnormal gait diagnosis and rehabilitation: a pilot study for Parkinson disease patients

Conventional gait rehabilitation treatment does not provide quantitative information on abnormal gait kinematics, and the match of the intervention strategy to the underlying clinical presentation may be limited by clinical expertise and experience. Also the effect of rehabilitation treatment may be reduced as the rehabilitation treatment is achieved only in a clinical setting. In this paper, a mobile gait monitoring system (MGMS) is proposed for the diagnosis of abnormal gait and rehabilitation. The proposed MGMS consists of Smart Shoes and a microsignal processor with a touch screen display. It monitors patients' gait by observing the ground reaction force (GRF) and the center of GRF, and analyzes the gait abnormality. Since visual feedback about patients' GRFs and normal GRF patterns are provided by the MGMS, patients can practice the rehabilitation treatment by trying to follow the normal GRF patterns without restriction of time and place. The gait abnormality proposed in this paper is defined by the deviation between the patient's GRFs and normal GRF patterns, which are constructed as GRF bands. The effectiveness of the proposed gait analysis methods with the MGMS has been verified by preliminary trials with patients suffering from gait disorders.     

Prediction of ground reaction forces in level and incline/decline walking from a multistage analysis of plantar pressure data

Knowing the ground reaction forces (GRFs) during walking has various biomechanical applications in injury prevention, gait analysis, as well as prosthetic and footwear design. The current study presents a method for predicting the GRFs in level and incline/decline walking that may be used in various outdoor biomechanics studies geared towards the above applications. The method was developed to predict the complete set of GRFs at walking inclinations of 0°, ±5°, ±10°, ±15°, and ±20°. Plantar pressure insoles were used to obtain inclination-specific, linear regression models based on three periods of gait stance phase, and the model-determined GRFs were compared with those measured from a forceplate. The three periods were determined based on the observed shifting of load-bearing insole sensors from heel to forefoot during walking, i.e., heel-strike, midstance, and toe-off. Six subjects wearing minimalist shoes fitted with plantar pressure insoles containing 99 pressure sensors performed ten walking trials at each of the aforementioned inclinations on an adjustable ramp with an embedded forceplate. Data from contact of the instrumented shoes with the forceplate were used to create linear regressions to transform insole pressure data into a complete set of GRFs. The root mean square error (RMSE) over peak recorded values were on average 10%, 3%, 21% for level walking and 11%, 4%, 23% for ramp walking in the respective anteroposterior, vertical, and mediolateral directions. The multistage linear regression model developed in the current study may be an acceptable option for estimating GRFs during walking in various environments without the restraint of a forceplate.     

Contributions of muscles to mediolateral ground reaction force over a range of walking speeds

Impaired control of mediolateral body motion during walking is an important health concern. Developing treatments to improve mediolateral control is challenging, partly because the mechanisms by which muscles modulate mediolateral ground reaction force (and thereby modulate mediolateral acceleration of the body mass center) during unimpaired walking are poorly understood. To investigate this, we examined mediolateral ground reaction forces in eight unimpaired subjects walking at four speeds and determined the contributions of muscles, gravity, and velocity-related forces to the mediolateral ground reaction force by analyzing muscle-driven simulations of these subjects. During early stance (0-6% gait cycle), peak ground reaction force on the leading foot was directed laterally and increased significantly (p<0.05) with walking speed. During early single support (14-30% gait cycle), peak ground reaction force on the stance foot was directed medially and increased significantly (p<0.01) with speed. Muscles accounted for more than 92% of the mediolateral ground reaction force over all walking speeds, whereas gravity and velocity-related forces made relatively small contributions. Muscles coordinate mediolateral acceleration via an interplay between the medial ground reaction force contributed by the abductors and the lateral ground reaction forces contributed by the knee extensors, plantarflexors, and adductors. Our findings show how muscles that contribute to forward progression and body-weight support also modulate mediolateral acceleration of the body mass center while weight is transferred from one leg to another during double support.     

Kinematic adaptations to a variable stiffness shoe: mechanisms for reducing joint loading

A recently described variable-stiffness shoe has been shown to reduce the adduction moment and pain in patients with medial-compartment knee osteoarthritis. The mechanism associated with how this device modifies overall gait patterns to reduce the adduction moment is not well understood. Yet this information is important for applying load modifying intervention for the treatment of knee osteoarthritis. A principal component analysis (PCA) was used to test the hypothesis that there are differences in the frontal plane kinematics that are correlated with differences in the ground reaction forces (GRFs) and center of pressure (COP) for a variable-stiffness compared to a constant-stiffness control shoe. Eleven healthy adults were tested in a constant-stiffness control shoe and a variable-stiffness shoe while walking at self-selected speeds. The PCA was performed on trial vectors consisting of all kinematic, GRF and COP data. The projection of trial vectors onto the linear combination of four PCs showed there were significant differences between shoes. The interpretation of the PCs indicated an increase in the ankle eversion, knee abduction and adduction, decreases in the hip adduction and pelvic obliquity angles and reduced excursion of both the COP and peak medial-lateral GRFs for the variable-stiffness compared to the control shoe. The variable-stiffness shoe produced a unique dynamic change in the frontal plane motion of the ankle, hip and pelvis that contributed to changes in the GRF and COP and thus reduced the adduction moment at a critical instant during gait suggesting a different mechanism that was seen with fixed interventions (e.g. wedges).     

Can treadmill walking be used to assess propulsion generation?

Instrumented treadmills offer significant advantages for analysis of human locomotion, including recording consecutive steady-state gait cycles, precisely controlling walking speed, and avoiding force plate targeting. However, some studies of hemiparetic walking on a treadmill have suggested that the moving treadmill belt may fundamentally alter propulsion mechanics. Any differences in propulsion mechanics during treadmill walking would be problematic since recent studies assessing propulsion have provided fundamental insight into hemiparetic walking. The purpose of this study was to test the hypothesis that there would be no difference in the generation of anterior/posterior (A/P) propulsion by performing a carefully controlled comparison of the A/P ground reaction forces (GRFs) and impulses in healthy adults during treadmill and overground walking. Gait data were collected from eight subjects walking overground and on a treadmill with speed and cadence controlled. Peak negative and positive horizontal GRFs in early and late stance, respectively, were reduced by less than 5% of body weight (p<0.05) during treadmill walking compared to overground walking. The magnitude of the braking impulse was similarly lower (p<0.05) during treadmill walking, but no significant difference was found between propulsion impulses. While there were some subtle differences in A/P GRFs between overground and treadmill walking, these results suggest there is no fundamental difference in propulsion mechanics. We conclude that treadmill walking can be used to investigate propulsion generation in healthy and by implication clinical populations.     

Resonance-based oscillations could describe human gait mechanics under various loading conditions

The oscillatory behavior of the center of mass (CoM) and the corresponding ground reaction force (GRF) of human gait for various gait speeds can be accurately described in terms of resonance using a spring–mass bipedal model. Resonance is a mechanical phenomenon that reflects the maximum responsiveness and energetic efficiency of a system. To use resonance to describe human gait, we need to investigate whether resonant mechanics is a common property under multiple walking conditions. Body mass and leg stiffness are determinants of resonance; thus, in this study, we investigated the following questions: (1) whether the estimated leg stiffness increased with inertia, (2) whether a resonance-based CoM oscillation could be sustained during a change in the stiffness, and (3) whether these relationships were consistently observed for different walking speeds. Seven healthy young subjects participated in over-ground walking trials at three different gait speeds with and without a 25-kg backpack. We measured the GRFs and the joint kinematics using three force platforms and a motion capture system. The leg stiffness was incorporated using a stiffness parameter in a compliant bipedal model that best fitted the empirical GRF data. The results showed that the leg stiffness increased with the load such that the resonance-based oscillatory behavior of the CoM was maintained for a given gait speed. The results imply that the resonance-based oscillation of the CoM is a consistent gait property and that resonant mechanics may be useful for modeling human gait.     

Lower-extremity ground reaction forces in collegiate baseball pitchers

The purpose of this study was to investigate ground reaction forces (GRF) in collegiate baseball pitchers and their relationship to pitching mechanics. Fourteen healthy collegiate baseball pitchers participated in this study. High-speed video and force plate data were collected for fastballs from each pitcher. The average ball speed was 35 ± 3 m/sec (78 ± 7 mph). Peak GRFs of 245 ± 20% body weight (BW) were generated in an anterior or braking direction to control descent. Horizontal GRFs tended to occur in a laterally directed fashion, reaching a peak of 45 ± 63% BW. The maximum vertical GRF averaged 202 ± 43% BW approximately 45 milliseconds after stride foot contact. A correlation between braking force and ball velocity was evident. Because of the downward inclination and rotation of the pitching motion, in addition to volume, shear forces may occur in the musculoskeletal tissues of the stride limb leading to many of the lower-extremity injuries seen in this athletic population.     

Determination of joint moments with instrumented force shoes in a variety of tasks

Ground reaction forces (GRFs) are often used in inverse dynamics analyses to determine joint loading. These GRFs are usually measured using force plates (FPs). As an alternative, instrumented force shoes (FSs) can be used, which have the advantage over FPs that they do not constrain foot placement. This study tested the FS system in one normal weight subject (77 kg) performing 19 different lifting, pushing and pulling and walking tasks. Kinematics were measured with an optoelectronic system and the GRFs and the positions of the centre of pressure (CoP) were synchronously measured with FPs and FSs. Differences between the outcomes of the two measurement systems (i.e. CoP and GRFs) and the resulting ankle and L5/S1 joint moments were determined at the instant of the peak GRF (DaPF). For most lifting and pushing and pulling tasks, the difference between the FP and FS measurements remained small: GRF DaPF remained below 3% body weight, CoP DaPF remained below 10 mm, ankle moment DaPF remained below 7% of the peak total ankle moment that occurred during normal walking and L5/S1 moment DaPF remained below 7% of the peak total L5/S1 moment that occurred during normal symmetric lifting. More substantial differences were only found in the maximal pushing tasks. For the walking tasks, peak vertical GRFs were somewhat underestimated. However, differences in ankle and L5/S1 moments remained small, i.e. DaPF below 7% of the peak total moment that occurred during normal walking.     

Experimental/analytical analysis of human locomotion using bondgraphs

A new vectorial bondgraph approach for modeling and simulation of human locomotion is introduced. The vectorial bondgraph is applied to an eight-segment gait model to derive the equations of motion for studying ground reaction forces (GRFs) and centers of pressure (COPs) in single and double support phases of ground and treadmill walking. A phase detection technique and accompanying transition equation is proposed with which the GRFs and COPs may be calculated for the transitions from double-to-single and single-to-double support phases. Good agreement is found between model predictions and experimental data obtained from force plate measurements. The bondgraph modeling approach is shown to be both informative and adaptable, in the sense that the model resembles the human body structure, and that modeled body segments can be easily added or removed.     

Effect of centerline-guided walking on gait characteristics in healthy subjects

One of the challenges in collecting ground reaction force (GRF) and moment data for gait analysis is to obtain “good hits” when the subject walks past the forceplates. We examined whether centerline-guided walking would significantly increase the chance of good hits and alter gait characteristics. Thirty-five healthy individuals (age: 37±13 yrs) walked on a walkway with five embedded forceplates at comfortable self-selected speeds under two conditions: (1) free walking and (2) walking along a centerline and avoiding stepping on it. Gait kinematics and GRF were collected using an 8-camera optoelectronic system and five forceplates, respectively. Surface electromyographic (EMG) activity of the rectus femoris, hamstring, gastrocnemius (GAS), and tibialis anterior (TA) were monitored bilaterally. The probability of good hits significantly increased with the centerline-guided walking (p=0.008). Repeated measures MANOVA and follow-up univariate tests revealed no significant differences between the two conditions in any of the spatiotemporal parameters except for a significant increase in step width with centerline walking (p<0.001). Centerline guiding significantly increased peak mediolateral GRF (p<0.001) and hip adduction/abduction and ankle internal/external rotation ranges of motion (p<0.01). In addition, the average EMG activity in GAS and TA during the stance phase significantly increased with the centerline walking (p<0.001). In general, the centerline walking tended to impact women more than men. Centerline-guided walking increases the chance of good hits but biomechanical characteristics of gait in the frontal and transverse planes and EMG activity should be interpreted with caution, especially in women.     

An actuated dissipative spring-mass walking model: Predicting human-like ground reaction forces and the effects of model parameters

Simple models are widely used to understand the mechanics of human walking. The optimization-based minimal biped model and spring-loaded-inverted-pendulum (SLIP) model are two popular models that can achieve human-like walking patterns. However, ground reaction forces (GRF) from these two models still deviate from experimental data. In this paper, we proposed an actuated dissipative spring-mass model by integrating these two models to realize more human-like GRF patterns. We first explored the function of stiffness, damping, and weights of both energy cost and force cost in the objective function and found that these parameters have distinctly different influences on the optimized gait and GRF profiles. The stiffness and objective weight affect the number and size of peaks in the vertical GRF and stance time. The damping changes the relative size of the peaks but has little influence on stance time. Based on these observations, these parameters were manually tuned at three different speeds to approach experimentally measured vertical GRF and the highest correlation coefficient can reach 0.983. These results indicate that the stiffness, damping, and proper objective functions are all important factors in achieving human-like motion for this simple walking model. These findings can facilitate the understanding of human walking dynamics and may be applied in future biped models.     

Estimates of muscle function in human gait depend on how foot-ground contact is modelled

Computational analyses of leg-muscle function in human locomotion commonly assume that contact between the foot and the ground occurs at discrete points on the sole of the foot. Kinematic constraints acting at these contact points restrict the motion of the foot and, therefore, alter model calculations of muscle function. The aim of this study was to evaluate how predictions of muscle function obtained from musculoskeletal models are influenced by the model used to simulate ground contact. Both single- and multiple-point contact models were evaluated. Muscle function during walking and running was determined by quantifying the contributions of individual muscles to the vertical, fore-aft and mediolateral components of the ground reaction force (GRF). The results showed that two factors--the number of foot-ground contact points assumed in the model and the type of kinematic constraint enforced at each point--affect the model predictions of muscle coordination. Whereas single- and multiple-point contact models produced similar predictions of muscle function in the sagittal plane, inconsistent results were obtained in the mediolateral direction. Kinematic constraints applied in the sagittal plane altered the model predictions of muscle contributions to the vertical and fore-aft GRFs, while constraints applied in the frontal plane altered the calculations of muscle contributions to the mediolateral GRF. The results illustrate the sensitivity of calculations of muscle coordination to the model used to simulate foot-ground contact.     

Control of ground reaction forces by hindlimb muscles during cat locomotion

It has been proposed that biarticular muscles are primarily responsible for the control of the direction of external forces, as their activation is closely related and highly sensitive to the direction of external forces. This functional role for biarticular muscles has been supported qualitatively by experimental evidence, but has never been tested quantitatively for lack of a mathematical/mechanical formulation of this theory and the difficulty of measuring individual muscle forces during voluntary movements. The purposes of this study were: (1) to define rules for muscular coordination based on the control of external forces; (2) to develop a model of the cat hindlimb that allows for the calculation of the magnitude and direction of the ground reaction forces (GRFs) produced by individual hindlimb muscles; and (3) to test if the coordination of mono- and biarticular cat hindlimb muscles is related to the control of the resultant GRF. We measured the GRF, hindlimb kinematics, selected muscle forces and activations during cat locomotion. Then, the measured muscle forces were used as input to the hindlimb model to compute the muscle-induced GRF. We assume that if activation (and possibly force) increased as the muscle-induced component of GRF approximated the resultant GRF, then that muscle was used by the central nervous system (CNS) to help control the direction of the external GRF. During cat walking, medial gastrocnemius (MG) and plantaris (PL) forces increased with increasing proximity to the GRF, while soleus (SOL) forces and vastus lateralis (VL) activations did not. SOL and VL activation were most strongly related to the vertical and parallel (braking/accelerating) component of the GRF, respectively. We concluded from these results that MG and PL are primarily responsible for the control of the direction of the GRF, while SOL primarily functions as an anti-gravity muscle, and VL as an acceleration/deceleration muscle.     

Biomechanical characteristics of adults walking in shallow water and on land.

Although water environment has been employed for different physical activities, there is little available information regarding the biomechanical characteristics of walking in shallow water. In the present study, we investigated the kinematics, ground reaction forces (GRF), and electromyographic (EMG) activation patterns of eight selected muscles of adults walking in shallow water and on land. Ten healthy adults were videotaped while walking at self-selected comfortable speeds on land and in water (at the Xiphoid process level). In both conditions there was a force plate embedded in the middle of each walkway to register the GRF components. Reflective markers were placed over main anatomical landmarks and they were digitalized later to obtain stride characteristics and joint angle information. In general, walking in water was different to walking on land in many aspects and these differences were attributed to the drag force, the apparent body weight reduction, and the lower comfortable speed during walking in shallow water. The joint range of motions (ROM) were not different, the segment ROM, magnitudes of GRF components, impact force, and impulse were different between the two conditions. The present results will contribute to a better understanding of this activity in the context of training and rehabilitation.     

Estimates of muscle function in human gait depend on how foot-ground contact is modelled

Computational analyses of leg-muscle function in human locomotion commonly assume that contact between the foot and the ground occurs at discrete points on the sole of the foot. Kinematic constraints acting at these contact points restrict the motion of the foot and, therefore, alter model calculations of muscle function. The aim of this study was to evaluate how predictions of muscle function obtained from musculoskeletal models are influenced by the model used to simulate ground contact. Both single- and multiple-point contact models were evaluated. Muscle function during walking and running was determined by quantifying the contributions of individual muscles to the vertical, fore-aft and mediolateral components of the ground reaction force (GRF). The results showed that two factors – the number of foot-ground contact points assumed in the model and the type of kinematic constraint enforced at each point – affect the model predictions of muscle coordination. Whereas single- and multiple-point contact models produced similar predictions of muscle function in the sagittal plane, inconsistent results were obtained in the mediolateral direction. Kinematic constraints applied in the sagittal plane altered the model predictions of muscle contributions to the vertical and fore-aft GRFs, while constraints applied in the frontal plane altered the calculations of muscle contributions to the mediolateral GRF. The results illustrate the sensitivity of calculations of muscle coordination to the model used to simulate foot-ground contact.     

Functional assessment in the rat by ground reaction forces

Although the rat sciatic nerve model is used extensively in the investigation of repair techniques, and a variety of evaluation methods utilized to assess the results, a means to measure directly and accurately the return of function in these animals is absent. Histologic, histomorphometric, and electrophysiologic methods can be reliable indicators of nerve regeneration but do not correlate to functional recovery. The purposes of this study were to develop apparatus to continuously measure ground reaction forces (GRF) and use GRF parameters in the assessment of gait parameters in normal rats preoperatively and following peripheral nerve severance and repair. Three neurorrhaphy methods: direct sciatic nerve repair, direct tibial nerve repair and double sciatic nerve repair simulating autograft, as well as a non-repaired tibial nerve transection were evaluated. The testing apparatus was designed to measure the spontaneous and voluntary effort of the rat with objective data. Three orthogonal components - vertical, craniocaudal (braking and propulsion), and mediolateral - of the ground reaction force were measured. Preoperative data showed that vertical forces were comparable among the four limbs but propulsion and braking forces displayed significant differences. At 12 weeks, functional recovery was most evident in the direct tibial nerve repair group and absent in the non-repaired tibial defect group. Direct sciatic nerve repairs and sciatic nerve grafts resulted in lesser degrees of improvement. Results indicated that the propulsive force is the optimal GRF parameter for evaluating recovery of useful function.