A computational method for reliable gait event detection and abnormality detection for feedback in rehabilitation

In this paper, a gait event detection algorithm is presented that uses computer intelligence (fuzzy logic) to identify seven gait phases in walking gait. Two inertial measurement units and four force-sensitive resistors were used to obtain knee angle and foot pressure patterns, respectively. Fuzzy logic is used to address the complexity in distinguishing gait phases based on discrete events. A novel application of the seven-dimensional vector analysis method to estimate the amount of abnormalities detected was also investigated based on the two gait parameters. Experiments were carried out to validate the application of the two proposed algorithms to provide accurate feedback in rehabilitation. The algorithm responses were tested for two cases, normal and abnormal gait. The large amount of data required for reliable gait-phase detection necessitate the utilisation of computer methods to store and manage the data. Therefore, a database management system and an interactive graphical user interface were developed for the utilisation of the overall system in a clinical environment.     

A computational method for reliable gait event detection and abnormality detection for feedback in rehabilitation

In this paper, a gait event detection algorithm is presented that uses computer intelligence (fuzzy logic) to identify seven gait phases in walking gait. Two inertial measurement units and four force-sensitive resistors were used to obtain knee angle and foot pressure patterns, respectively. Fuzzy logic is used to address the complexity in distinguishing gait phases based on discrete events. A novel application of the seven-dimensional vector analysis method to estimate the amount of abnormalities detected was also investigated based on the two gait parameters. Experiments were carried out to validate the application of the two proposed algorithms to provide accurate feedback in rehabilitation. The algorithm responses were tested for two cases, normal and abnormal gait. The large amount of data required for reliable gait-phase detection necessitate the utilisation of computer methods to store and manage the data. Therefore, a database management system and an interactive graphical user interface were developed for the utilisation of the overall system in a clinical environment.     

Normal hip joint contact pressure distribution in single-leg standing--effect of gender and anatomic parameters.

A practical and easy-to-use analysis technique that can study the patient's hip joint contact force/pressure distribution would be useful to assess the effect of abnormal biomechanical conditions and anatomical deformities on joint contact stress for treatment planning purpose. This technique can also help to establish the normative database on hip joint contact pressure distribution in men and women in different age groups. Twelve anatomic parameters and seven biomechanical parameters of the hip joint in a normal population (41 females, 15 males) were calculated. The inter-parameter correlations were investigated. The pressure distribution in the hip joint was calculated using a three-dimensional discrete element analysis (DEA) technique. The 3D contact geometry of the hip joint was estimated from a 2D radiograph by assuming that the femoral head and the acetabular surface were spherical in shape. The head-trochanter ratio (HT), femoral head radius, pelvic height, the joint contact area, the normalized peak contact pressure, abductor force, and the joint contact force were significantly different between men and women. The normalized peak contact pressure was correlated both with acetabular coverage and head-trochanter ratio. Change of abductor force direction within normal variation did not affect the joint peak contact pressure. However, in simulated dysplastic conditions when the CE angle is small or negative, abductor muscle direction becomes very sensitive in joint contact pressure estimation. The models and the results presented can be used as the reference base in computer simulation for preoperative planning in pelvic or femoral osteotomy.     

Foot pressure patterns during gait

Peak pressure and temporal parameters of foot function were determined in 21 men and 11 women; few differences between men and women in any of the parameters were observed, either when walking barefoot or when wearing shoes. However, significant differences were observed when barefoot and shod walking were compared. The main influence of shoes appears to be in modifying the behaviour of the forefoot, by changing the pressure distribution across the metatarsal heads and increasing the contact times for the toes. The implications of such changes are discussed. Intersubject variability in the pattern and magnitude of the peak pressure distributions under the foot, which appear to be consistent in both the short and long term, may prove to be of importance in the clinical environment, by providing an individual baseline from which change can be measured. A new measure, the pressure-time integral, could be a more valuable single measure than either the peak pressure or the temporal parameters.     

Comparison between sensitive and nonsensitive free flaps in reconstruction of the heel and plantar area

In this study, 12 cases of reconstruction of the heel and plantar area since 1982 are reviewed. Six nonsensate muscle free flaps and six sensate fasciocutaneous flaps were used, respectively. Categories assessed were the time interval for return to daily living activities, sensation to light touch, pinprick, Semmes-Weinstein monofilament test of the reconstructed area for sensory evaluation; and results of pedograms (maximal pressure, pressure distribution, and total contact area of the plantar surface). Follow-up periods were between 2 and 14 years, with an average of 6 years. Better sensory results and early return to daily living activities were observed in the sensate flap group, but the defects were smaller in this group. Despite the slightly longer time to return to daily living activities and worse sensory results, long-term follow-up showed that patients with nonsensate flaps had no difficulty in performing living activities if they continued to be careful and to use some kind of protective shoes. The results of the pedogram analyses were similar between the two groups with regard to total contact area of the reconstructed foot in relation to the healthy foot. Pressure values of the reconstructed areas in sensate flaps were found to be close to pressure values in the same weight areas of the normal foot. The differences between pressure values of the sensate and nonsensate flaps were statistically significant (p < 0.001). Therefore, in reconstruction of the weight-bearing surface of the foot, each case should be evaluated individually. The reconstructive method should be chosen according to the location and soft-tissue requirements of the defect.     

Dynamic plantar pressure distribution during terrestrial locomotion of bonobos (Pan paniscus)

We collected high-resolution plantar pressure distributions of seven bonobos during terrestrial bipedal and quadrupedal locomotion (N = 146). Functional foot length, degree of hallux abduction, and total contact time were determined, and plots, showing pressure as a function of time for six different foot regions, were generated. We also studied five adult humans for comparison (N = 13). Both locomotion types of the bonobo show a large variation in plantar pressure distributions, which could be due to the interference of instantaneous behavior with locomotion and differences in walking speed and body dimensions. The heel and the lateral midfoot typically touch down simultaneously at initial ground contact in bipedal and quadrupedal walking of bonobos, in contrast with the typical heel-strike of human bipedalism. The center of pressure follows a curved course during quadrupedalism, as a consequence of the medial weight transfer during mid-stance. Bipedal locomotion of bonobos is characterized by a more plantar positioning of the feet and by a shorter contact time than during quadrupedal walking, according to a smaller stride and step length at a higher frequency. We observed a varus position of the foot with an abducted hallux, which likely possesses an important sustaining and stabilizing function during terrestrial locomotion.     

Effect of heel height on in-shoe localized triaxial stresses

Abnormal and excessive plantar pressure and shear are potential risk factors for high-heeled related foot problems, such as forefoot pain, hallux valgus deformity and calluses. Plantar shear stresses could be of particular importance with an inclined supporting surface of high-heeled shoe. This study aimed to investigate the contact pressures and shear stresses simultaneously between plantar foot and high-heeled shoe over five major weightbearing regions: hallux, heel, first, second and fourth metatarsal heads, using in-shoe triaxial force transducers. During both standing and walking, peak pressure and shear stress shifted from the lateral to the medial forefoot as the heel height increased from 30 to 70mm. Heel height elevation had a greater influence on peak shear than peak pressure. The increase in peak shear was up to 119% during walking, which was about five times that of peak pressure. With increasing heel height, peak posterolateral shear over the hallux at midstance increased, whereas peak pressure at push-off decreased. The increased posterolateral shear could be a contributing factor to hallux deformity. It was found that there were differences in the location and time of occurrence between in-shoe peak pressure and peak shear. In addition, there were significant differences in time of occurrence for the double-peak loading pattern between the resultant horizontal ground reaction force peaks and in-shoe localized peak shears. The abnormal and drastic increase of in-shoe shear stresses might be a critical risk factor for shoe-related foot disorders. In-shoe triaxial stresses should therefore be considered to help in designing proper footwear.     

Ultrasonic motion analysis system--measurement of temporal and spatial gait parameters

The duration of stance and swing phase and step and stride length are important parameters in human gait. In this technical note a low-cost ultrasonic motion analysis system is described that is capable of measuring these temporal and spatial parameters while subjects walk on the floor. By using the propagation delay of sound when transmitted in air, this system is able to record the position of the subjects' feet. A small ultrasonic receiver is attached to both shoes of the subject while a transmitter is placed stationary on the floor. Four healthy subjects were used to test the device. Subtracting positions of the foot with zero velocity yielded step and stride length. The duration of stance and swing phase was calculated from heel-strike and toe-off. Comparison with data obtained from foot contact switches showed that applying two relative thresholds to the speed graph of the foot could reliably generate heel-strike and toe-off. Although the device is tested on healthy subjects in this study, it promises to be extremely valuable in examining pathological gait. When gait is asymmetrical, walking speed is not constant or when patients do not completely lift their feet, most existing devices will fail to correctly assess the proper gait parameters. Our device does not have this shortcoming and it will accurately demonstrate asymmetries and variations in the patient's gait. As an example, the recording of a left hemiplegic patient is presented in the discussion.     

Assessment of the influence of foot orthoses in the hip loading conditions during walking: a single case study

Despite their large clinical application, the understanding of the effects of foot orthoses on the lower limb kinematics and kinetics is limited. In this context, we propose an advanced musculoskeletal model to assess the influence of foot orthoses in the loading conditions within an osteoarthritic hip joint during gait. Experimental data are collected for a single pathological subject presenting a coxarthrosis (with and without orthoses), and a healthy subject during walking. An inverse dynamic approach coupled with an optimisation method evaluates the forces developed by 14 muscles and the hip contact reaction force. Contact reaction and muscular force magnitudes are closed whether the patient is walking with or without foot orthoses. Nevertheless, contact reaction amplitudes and orientations show differences in relation to those calculated for the healthy subject. The results obtained allow us to formulate some assumptions concerning the causes of coxarthrosis evolution and treatment.     

Energy-conserving impact algorithm for the heel-strike phase of gait

Significant ground reaction forces exceeding body weight occur during the heel-strike phase of gait. The standard methods of analytical dynamics used to solve the impact problem do not accommodate well the heel-strike collision due to the persistent contact at the front foot and presence of contact at the back foot. These methods can cause a non-physical energy gain on the order of the total kinetic energy of the system at impact. Additionally, these standard techniques do not quantify the contact force, but the impulse over the impact. We present an energy-conserving impact algorithm based on the penalty method to solve for the ground reaction forces during gait. The rigid body assumptions are relaxed and the bodies are allowed to penetrate one another to a small degree. Associated with the deformation is a potential, from which the contact forces are derived. The empirical coefficient-of-restitution used in the standard approaches is replaced by two parameters to characterize the stiffness and the damping of the materials. We solve two simple heel-strike models to illustrate the shortcomings of a standard approach and the suitability of the proposed method for use with gait.     

Influence of the posterior tibial tendon on the medial arch of the foot: an in vitro kinetic and kinematic study]

INTRODUCTION: The respective contributions of the active and passive structures of the foot to the stability of the medical arch were investigated using an in vitro kinetic and kinematic model. The effect of the tibialis posterior tendon on foot and ankle movements, and plantar pressure distribution of the foot were tested in a cadaveric human foot. METHOD: The stance phase from heel-contact to toe-off of normal walking gait and after tibialis posterior tendon rupture was simulated in eight roentenographically normal human feet (age 66 +/- 19 years, males). Ground reaction force and tibial inclination was simulated by means of a tilting angle and force-controlled translation stage. Plantar pressure was measured using a pressure-measuring platform. The force developed by the flexors and extensor muscles of the foot were simulated via cables attached to 7 force-controlled hydraulic cylinders. Tibial rotation was produced by an electric servo-motor, and foot movements measured with an ultrasonic analysis system. RESULTS: The model was verified against the plantar distribution and kinematics of healthy subjects measured during normal gait. Tibialis posterior deficit did not result in any detectable changes in pressure or force-time integral in the medial regions of the foot--a common sign of flat foot (pressure: midfoot 0.2 < or = 0.9; medial forefoot 0.5 < or = p < or = 0.9; hallux 0.5 < or = p < or = 0.9; force-time integral: midfoot p = 0-871; medial forefoot p = 0.632; hallux p = 0.068). Only small tendential changes in the kinematics of the talus and calcaneus were observed in dorsiflexion (0-58 sec; talus 0.1 < or = p < or = 0.6; calcaneus 0.4 < or = p < or = 0.06) and eversion (talus: 0-60 sec. 0.1 < or = p < or = 0.6; calcaneus: 37-60 sec. 0.2 < or = p < or = 0.7). CONCLUSION: The results of this in vitro study show that defective tibialis posterior alone does not produce significant changes in the kinetics or kinematics of the stance phase of normal gait. This suggests that the development of flat foot observed in degeneration of the tibialis posterior tendon occurs only after fatigue of the passive structures of the foot.     

Genetics of the human electroencephalogram (EEG) and event-related brain potentials (ERPs): a review

Twin and family studies of normal variation in the human electroencephalogram (EEG) and event related potentials (ERPs) are reviewed. Most of these studies are characterized by small sample sizes. However, by summarizing these studies in one paper, we may be able to gain some insight into the genetic influences on individual differences in central nervous system functioning that may mediate genetically determined differences in behavior. It is clear that most EEG parameters are to a large extent genetically determined. The results for ERPs are based on a much smaller number of studies and suggest medium to large heritability.     

Age-related,interindividual, and right/left differences in anterior-posterior foot pressure ratio in preschool children

Background

This study aimed to examine age-related, interindividual, and right/left differences in anterior-posterior foot pressure ratio in 764 preschool children (364 boys and 400 girls) aged 3.5-6.5 years.

Methods

Subjects maintained an upright standing posture for 10 seconds on the Footview Clinic, an instrument designed to calculate the anterior-posterior foot pressure ratio. The ratio of anterior foot pressure in each subject’s right and left feet was selected as a variable, and the mean of a 10 s measurement was used for analysis.

Results

The ratio of anterior foot pressure was significantly larger in the right foot than in the left foot. With regard to age, the ratio of anterior foot pressure was significantly larger in children aged over 4.5 years than in children aged 3.5 years. It was also larger in children aged 6 and 6.5 years than in children aged 4 years. Interindividual differences in variables were large, and coefficients of variance were highest in children aged 3.5 years and lowest in children aged 6.5 years.

Conclusions

In conclusion, anterior foot pressure increases with age in preschool children. Interindividual differences in anterior foot pressure are large and tend to decrease with age. Furthermore, the anterior foot pressure is slightly higher in the right foot than in the left foot. These results will be useful for various studies, such as examining relationships between the anterior-posterior foot pressure ratio and factors, such as untouched toes, physical fitness, and level of exercise.     

Identification of cross-sectional parameters of lateral meniscal allografts that predict tibial contact pressure in human cadaveric knees

To guide the development of improved procedures for selecting meniscal allografts, the objective of this study was to identify which cross-sectional parameters of a lateral meniscal allograft predict the contact pressure of the articular surface of the tibia. To meet the objective, the contact pressure of the articular surface of the tibia was measured with a lateral meniscal autograft and a lateral meniscal allograft using pressure sensitive film in 15 fresh-frozen human cadaveric knees. Allografts were matched only in transverse dimensions to the autograft but not in cross-sectional dimensions. Knees were loaded to 1200 N in compression at flexion angles of 0, 15, 30 and 45 degrees using a load application system that allowed unconstrained motion in the remaining degrees of freedom. Five cross-sectional parameters for both of the grafts in each of the anterior, middle, and posterior regions were derived from measurements obtained using a laser-based non-contacting three-dimensional coordinate digitizing system (3-DCDS) (Haut et al., J. Orthop Res, 2000). Five contact variables (i.e. the maximum pressure, mean pressure, contact area, and anterior-posterior and medial-lateral locations of the centroid of contact area) were determined from the pressure sensitive film. When each allograft was paired with the corresponding autograft, the root mean squared percent differences for the cross-sectional parameters ranged from a minimum of 28% for the width of the posterior region to 572% for the height of the posterior region. The root mean squared percent differences between the contact variables for paired grafts were 29% for the maximum pressure, 19% for the mean pressure, and 24% for the contact area. Differences in the cross-sectional parameters between the grafts were related to differences in the contact variables using regression analysis. Difference in the width was most often a predictor variable in the regression models with R2 values > or = 0.45. Differences in all of the four remaining cross-sectional parameters were also important predictor variables. Because failure to match cross-sectional parameters causes substantial difference in contact variables between an allograft and autograft and because cross-sectional parameters predict the contact pressure on the tibial plateau, protocols used to prospectively select allografts should concentrate on matching cross-sectional parameters and particularly the width to those of the original meniscus.     

Analysis of the intra-individual differences of the joint surfaces of the calcaneus

Patients with calcaneus fractures experience considerable interferences with daily living activities. The quality of anatomical reconstruction is important because of its influence on functional outcome. The aim of this study was to develop an automatic algorithm based on computer tomographic (CT) images to quantify the integrity of calcaneal joint surfaces. Validation of this algorithm was done by assessing intra-individual variations of characteristic joint parameters. Bilateral hind foot CT data of 12 subjects were manually segmented, and 3D models from the calcaneus, talus and cuboid were generated. These models were implemented in a custom-made software to analyse the area, 3D orientations and bone distance of the joint surfaces of the calcaneus. Three joints were detected, and the calculated parameters were compared between right and left hind foot by the evaluation of the directional asymmetry (%DA). The results were statistically analysed with a paired t-test. The median of area (5–7 %DA) of the joint surfaces and the distance between two articulating surfaces (8–9 %DA) showed the greatest intra-individual differences. Median differences in 3D orientation were comparatively low (1–2 %DA). None of these differences was statistically significant. Inter-individual variations among subjects were several magnitudes larger than intra-individual differences. The presented computational tool provides 3D joint-specific parameters of the calcaneus, which enable to describe their respective joint integrity. The results show that only small intra-individual differences within the anatomy exist. Surgical treatment should take place with the aid of CT data from the contralateral side. Thus, a good restoration of the anatomy may be reached. The computational tool assesses the quality of reduction, and may be helpful to evaluate the outcome and quality of operative treatment based on the calculated joint-specific parameters of joint reconstructions in the hind foot.     

Scaling of plantar pressures in mammals

The interaction of the limbs with the substrate can teach a lot about an animal's gait mechanics. Unlike ground-reaction forces, plantar pressure distributions are rarely studied in animals, but they may provide more detailed information about the loading patterns and locomotor function of specific anatomical structures. With this study, we aim to describe pressures for a large and diverse sample of mammalian species, focusing on scaling effects. We collected dynamic plantar pressure distributions during voluntary walking in 28 mammal species. A dynamic classification of foot use was made, which distinguished between plantiportal, digitiportal and unguliportal animals. Analysis focused on scaling effects of peak pressures, peak forces and foot contact areas. Peak pressure for the complete mammal sample was found to scale to (mass)^1/2, higher than predicted assuming geometric similarity, and we found no difference between the different types of foot use. Only the scaling of peak force is dependent on the dynamic foot use. We conclude that plantar peak pressure rises faster with mass than expected, regardless of the type of foot use, and scales higher than in limb bones. These results might explain some anatomical and behavioural adaptations in graviportal animals.     

The Relationships between Foot Arch Volumes and Dynamic Plantar Pressure during Midstance of Walking in Preschool Children

Objectives

The purpose of this study was to examine the correlation between the foot arch volume measured from static positions and the plantar pressure distribution during walking.

Methods

A total of 27 children, two to six years of age, were included in this study. Measurements of static foot posture were obtained, including navicular height and foot arch volume in sitting and standing positions. Plantar pressure, force and contact areas under ten different regions of the foot were obtained during walking.

Results

The foot arch index was correlated (r = 0.32) with the pressure difference under the midfoot during the foot flat phase. The navicular heights and foot arch volumes in sitting and standing positions were correlated with the mean forces and pressures under the first (r = −0.296∼−0.355) and second metatarsals (r = −0.335∼−0.504) and midfoot (r = −0.331∼−0.496) during the stance phase of walking. The contact areas under the foot were correlated with the foot arch parameters, except for the area under the midfoot.

Conclusions

The foot arch index measured in a static position could be a functional index to predict the dynamic foot functions when walking. The foot arch is a factor which will influence the pressure distribution under the foot. Children with a lower foot arch demonstrated higher mean pressure and force under the medial forefoot and midfoot, and lower contact areas under the foot, except for the midfoot region. Therefore, children with flatfoot may shift their body weight to a more medial foot position when walking, and could be at a higher risk of soft tissue injury in this area.     

Three-dimensional finite element analysis of the foot during standing--a material sensitivity study

Information on the internal stresses/strains in the human foot and the pressure distribution at the plantar support interface under loading is useful in enhancing knowledge on the biomechanics of the ankle-foot complex. While techniques for plantar pressure measurements are well established, direct measurement of the internal stresses/strains is difficult. A three-dimensional (3D) finite element model of the human foot and ankle was developed using the actual geometry of the foot skeleton and soft tissues, which were obtained from 3D reconstruction of MR images. Except the phalanges that were fused, the interaction among the metatarsals, cuneiforms, cuboid, navicular, talus, calcaneus, tibia and fibula were defined as contact surfaces, which allow relative articulating movement. The plantar fascia and 72 major ligaments were simulated using tension-only truss elements by connecting the corresponding attachment points on the bone surfaces. The bony and ligamentous structures were embedded in a volume of soft tissues. The encapsulated soft tissue was defined as hyperelastic, while the bony and ligamentous structures were assumed to be linearly elastic. The effects of soft tissue stiffening on the stress distribution of the plantar surface and bony structures during balanced standing were investigated. Increases of soft tissue stiffness from 2 and up to 5 times the normal values were used to approximate the pathologically stiffened tissue behaviour with increasing stages of diabetic neuropathy. The results showed that a five-fold increase in soft tissue stiffness led to about 35% and 33% increase in the peak plantar pressure at the forefoot and rearfoot regions, respectively. This corresponded to about 47% decrease in the total contact area between the plantar foot and the horizontal support surface. Peak bone stress was found at the third metatarsal in all calculated cases with a minimal increase of about 7% with soft tissue stiffening.     

A technique for normalizing centre of pressure paths

Centre of pressure paths may be expected to provide useful information in the analysis of gait. In this paper, a technique for normalizing and averaging centre of pressure paths is presented. It utilizes force data obtained from force platform measurements together with information about the placement of the foot on the force platform. This placement is described by specifying the location of the 'centre' and 'midline' of the foot which are defined for a footprint. The coordinate system associated with a footprint is defined and centre of pressure locations are determined in relation to this system. A comparison of the centre of pressure paths obtained using this method and the method proposed by Cavanagh [J. Biomechanics 11, 487-491 (1978)] shows significant differences. It is concluded that the selection of an appropriate technique to obtain centre of pressure paths is important.     

Hip contact stress during normal and staircase walking: the influence of acetabular anteversion angle and lateral coverage of the acetabulum

Hip contact stress is considered to be an important biomechanical factor related to development of coxarthrosis. The effect of the lateral coverage of the acetabulum on the hip contact stress has been demonstrated in several studies of hip dysplasia, whereas the effect of the anterior anteversion remains unclear. Therefore, the joint hip contact stress during normal level walking and staircase walking, in normal and dysplastic hips, for small and large acetabular anteversion angle was computed. For small acetabular anteversion angle, the hip contact stress is slightly increased (less than 15%) in staircase walking when compared with normal walking. In hips with large angle of acetabular anteversion, walking downstairs significantly increases the maximal peak contact stress (70% in normal hips and 115% in dysplastic hips) whereas walking upstairs decreases the peak contact stress (4% in normal hips and 34% in dysplastic hips) in comparison to normal walking. Based on the presented results, we suggest that the acetabular anteversion should be considered in biomechanical evaluation of the hips, especially when the lateral coverage of the acetabulum is small.