Sagittal and transverse ankle angle coupling can influence prosthetic socket transverse plane moments.

Introduction: The intact foot and ankle comprise a complex set of joints that allow rotation in multiple planes of motion. Some of these motions are coupled, meaning rotation in one plane induces motion in another. One such coupling is between the sagittal and transverse planes. For every step, plantar- and dorsi-flexion motion is coupled with external and internal rotation of the shank relative to the foot, respectively. There is no prosthetic foot available for prescription that mimics this natural coupling. The purpose of this study was to determine if a sagittal:transverse ankle angle coupling ratio exists that minimizes the peak transverse plane moment during prosthetic limb stance. Methods: A novel, torsionally active prosthesis (TAP) was used to couple sagittal and transverse plane motions using a 60-watt motor. An embedded controller generated transverse plane rotation trajectories proportional to sagittal plane ankle angles corresponding to sagittal:transverse coupling ratios of 1:0 (rigid coupling analogous to the standard-of-care), 6:1, 4:1, 3:1, and 2:1. Individuals with unilateral transtibial amputation were block randomized to walk in a straight line and in both directions around a 2 m circle at their self-selected speed with the TAP set at randomized coupling ratios. The primary outcome was the peak transverse plane moment, normalized to body mass, during prosthetic limb stance. Secondary outcomes included gait biomechanic metrics and a measure of satisfaction. Results: Eleven individuals with unilateral transtibial amputations participated in the study. The 6:1 coupling ratio resulted in reduced peak transverse plane moments in pairwise comparisons with 3:1 and 2:1 coupling ratios while walking in a straight line and with the prosthesis on the outside of the circle (p < .05). Coupling ratio had no effect on gait biomechanic metrics or satisfaction. Discussion: The general pattern of results suggests a quadratic relationship between the peak transverse plane moment and coupling ratio with a minimum at the 6:1 coupling ratio. The coupling ratio did not appear to adversely affect propulsion or body support. Subjects indicated they found all coupling ratios to be comfortable. While a mechatronic prosthesis like the TAP may have limited commercial potential, our future work includes testing a robust, passive prosthetic foot with a fixed coupling ratio.

Influence of Walking Over Unexpected Uneven Terrain on Joint Loading for Individuals With Transtibial Amputation.

Individuals with transtibial amputation (TTA) experience asymmetric lower-limb loading which can lead to joint pain and injuries. However, it is unclear how walking over unexpected uneven terrain affects their loading patterns. This study sought to use modeling and simulation to determine how peak joint contact forces and impulses change for individuals with unilateral TTA during an uneven step and subsequent recovery step and how those patterns compare to able-bodied individuals. We expected residual limb loading during the uneven step and intact limb loading during the recovery step would increase relative to flush walking. Further, individuals with TTA would experience larger loading increases compared to able-bodied individuals. Simulations of individuals with TTA showed during the uneven step, changes in joint loading occurred at all joints except the prosthetic ankle relative to flush walking. During the recovery step, intact limb joint loading increased in early stance relative to flush walking. Simulations of able-bodied individuals showed large increases in ankle joint loading for both surface conditions. Overall, increases in early stance knee joint loading were larger for those with TTA compared to able-bodied individuals during both steps. These results suggest that individuals with TTA experience altered joint loading patterns when stepping on uneven terrain. Future work should investigate whether an adapting ankle-foot prosthesis can mitigate these changes to reduce injury risk.

A Data-Driven and Personalized Stance Symmetry Controller for Robotic Ankle-Foot Prostheses: A Preliminary Investigation.

People with unilateral transtibial amputation generally exhibit asymmetric gait, likely due to inadequate prosthetic ankle function. This results in compensatory behavior, leading to long-term musculoskeletal impairments (e.g., osteoarthritis in the joints of the intact limb). Powered prostheses can better emulate biological ankles, however, control methods are over-reliant on non-disabled data, require extensive amounts of tuning by experts, and cannot adapt to each user's unique gait patterns. This work directly addresses all these limitations with a personalized and data-driven control strategy. Our controller uses a virtual setpoint trajectory within an impedance-inspired formula to adjust the dynamics of the robotic ankle-foot prosthesis as a function of stance phase. A single sensor measuring thigh motion is used to estimate the gait phase in real time. The virtual setpoint trajectory is modified via a data-driven iterative learning strategy aimed at optimizing ankle angle symmetry. The controller was experimentally evaluated on two people with transtibial amputation. The control scheme successfully increased ankle angle symmetry about the two limbs by 24.4% when compared to the passive condition. In addition, the symmetry controller significantly increased peak prosthetic ankle power output at push-off by 0.52 W/kg and significantly reduced biomechanical risk factors associated with osteoarthritis (i.e., knee and hip abduction moments) in the intact limb. This research demonstrates the benefits of personalized and data-driven symmetry controllers for robotic ankle-foot prostheses.

The Influence of Multiple Pregnancies on Gait Asymmetry: A Case Study.

Gait asymmetry is a predictor of fall risk and may contribute to increased falls during pregnancy. Previous work indicates that pregnant women experience asymmetric joint laxity and pelvic tilt during standing and asymmetric joint moments and angles during walking. How these changes translate to other measures of gait asymmetry remains unclear. Thus, the purpose of this case study was to determine the relationships between pregnancy progression, subsequent pregnancies, and gait asymmetry. Walking data were collected from an individual during 2 consecutive pregnancies during the second and third trimesters and 6 months postpartum of her first pregnancy and the first, second, and third trimesters and 6 months postpartum of her second pregnancy. Existing asymmetries in step length, anterior-posterior (AP) impulses, AP peak ground reaction forces, lateral impulses, and joint work systematically increased as her pregnancy progressed. These changes in asymmetry may be attributed to pelvic asymmetry, leading to asymmetric hip flexor and extensor length, or due to asymmetric plantar flexor strength, as suggested by her ankle work asymmetry. Relative to her first pregnancy, she had greater asymmetry in step length, step width, braking AP impulse, propulsive AP impulse, and peak braking AP ground reaction force during her second pregnancy, which may have resulted from increased joint laxity.

Using Deep Learning Models to Predict Prosthetic Ankle Torque.

Inverse dynamics from motion capture is the most common technique for acquiring biomechanical kinetic data. However, this method is time-intensive, limited to a gait laboratory setting, and requires a large array of reflective markers to be attached to the body. A practical alternative must be developed to provide biomechanical information to high-bandwidth prosthesis control systems to enable predictive controllers. In this study, we applied deep learning to build dynamical system models capable of accurately estimating and predicting prosthetic ankle torque from inverse dynamics using only six input signals. We performed a hyperparameter optimization protocol that automatically selected the model architectures and learning parameters that resulted in the most accurate predictions. We show that the trained deep neural networks predict ankle torques one sample into the future with an average RMSE of 0.04 ± 0.02 Nm/kg, corresponding to 2.9 ± 1.6% of the ankle torque's dynamic range. Comparatively, a manually derived analytical regression model predicted ankle torques with a RMSE of 0.35 ± 0.53 Nm/kg, corresponding to 26.6 ± 40.9% of the ankle torque's dynamic range. In addition, the deep neural networks predicted ankle torque values half a gait cycle into the future with an average decrease in performance of 1.7% of the ankle torque's dynamic range when compared to the one-sample-ahead prediction. This application of deep learning provides an avenue towards the development of predictive control systems for powered limbs aimed at optimizing prosthetic ankle torque.

Biomechanical responses of individuals with transtibial amputation stepping on a coronally uneven and unpredictable surface.

Coronally uneven surfaces are prevalent in natural and man-made terrain, such as holes or bumps in the ground, curbs, sidewalks, and driveways. These surfaces can be challenging to navigate, especially for individuals with lower limb amputations. This study examined the biomechanical response of individuals with unilateral transtibial amputation (TTA) taking a step on a coronally uneven surface while wearing their clinically prescribed prosthesis, compared to individuals without mobility impairments (controls). An instrumented walkway was used with the middle force plate positioned either flush or rotated ± 15˚ in the coronal plane and concealed (blinded). TTAs used greater hip abduction compared to controls across all conditions, but especially during blinded inversion. The recovery step width of TTAs was wider after blinded eversion and narrower after blinded inversion, but unchanged for controls. These results suggest TTAs may have decreased balance control on unexpected, uneven surfaces. Additionally, TTAs generated less positive prosthetic ankle joint work during blinded inversion and eversion, and less negative coronal hip joint work during blinded inversion compared to controls. These biomechanical responses could lead to increased energy expenditure on uneven terrain. Surface condition had no effect on the vertical center of mass for either group of participants. Finally, the TTAs and the control group generated similar vertical GRF impulses, suggesting the TTAs had sufficient body support despite differences in surface conditions. These results are important to consider for future prosthetic foot designs and rehabilitation strategies.

Using theoretical frameworks to examine fall history and associated prosthetic mobility in people with nondysvascular lower limb amputation.

BACKGROUND: Over a million people live with lower limb amputation (LLA) in the United States, and many of them will experience a fall in the next year. The aim of this study was to use existing theoretical frameworks in an attempt to organize the complex interactions of reported fall history and prosthetic mobility in community-ambulating people with LLA. METHODS: Self-reported fall rate and fall circumstances were recorded in a cross-section of people with unilateral LLA due to nondysvascular causes. Self-report and performance-based standardized outcome measures assessed prosthetic mobility and balance confidence. All variables were considered and appropriately placed within a proposed International Classification of Functioning, Disability, and Health framework while using a fall-type classification framework to classify fall circumstances. RESULTS: Information from 69 participants was analyzed. The reported fall rate was at 46%, with those with transfemoral amputation reporting significantly more falls than those with transtibial amputation ( P = 0.001). Tripping over an object was the most common cause (62.5%), and fallers reported significantly lower perceived prosthetic mobility than nonfallers ( P = 0.001). Despite reporting high levels of balance confidence, results indicate that all groups of fallers and nonfallers are at increased fall risk according to performance-based prosthetic mobility score cutoffs. CONCLUSIONS: Community-dwelling people with nondysvascular LLA are at increased fall risk. Classifying fall-related variables using theoretical frameworks provides a means to structure more informative fall risk surveys for people with LLA in an attempt to identify those at greater risk for falling and its potential detrimental effects.

Locomotor activities of individuals with lower-limb amputation.

BACKGROUND: Ambulatory individuals with lower-limb amputation perform a variety of locomotor activities, but the step count distribution of these activities is unknown. OBJECTIVE: To describe a novel method for activity monitoring and to use it to count steps taken while walking straight ahead on level ground, turning right and left, up and down stairs, and up and down ramps. STUDY DESIGN: This is an observational study. METHODS: A portable instrument to record leg motion was placed on or inside the prosthetic pylon of 10 individuals with unilateral transtibial amputations. Participants first walked a defined course in a hospital environment to train and validate a machine learning algorithm for classifying locomotor activity. Participants were then free to pursue their usual activities while data were continuously collected over 1-2 d. RESULTS: Overall classification accuracy was 97.5% ± 1.5%. When participants were free to walk about their home, work, and community environments, 82.8% of all steps were in a straight line, 9.0% were turning steps, 4.8% were steps on stairs, and 3.6% were steps on ramps. CONCLUSION: A novel activity monitoring method accurately classified the locomotion activities of individuals with lower-limb amputation. Nearly 1 in 5 of all steps taken involved turning or walking on stairs and ramps.

Biomechanical response to mediolateral foot-placement perturbations during walking.

Dynamic balance in the frontal plane requires active control, which is accomplished largely through control of mediolateral foot placement. Individuals without mobility impairments have the ability to compensate for variability in foot-placement to maintain their balance; however, it is unknown how individuals respond to unexpected mediolateral perturbations to their foot placement that alter their balance control. The purpose of this study was to identify the biomechanical responses of individuals without mobility impairments to medial and lateral foot-placement perturbations during walking. Three-dimensional body segment kinematic and ground reaction force data were collected from 15 participants at 1.0 m/s and their self-selected speed on an instrumented treadmill. Dynamic balance was assessed by analyzing whole-body angular momentum in the frontal plane. We hypothesized that participants would respond to the perturbations with a combination of a lateral ankle strategy, hip adduction strategy and/or ankle push-off strategy to restore their balance. Overall, the medial perturbations adversely affected dynamic balance while lateral perturbations had little effect. Individuals responded to medial (lateral) perturbations with an increased (decreased) ankle inversion moment, which correlated to lateral (medial) shifts in their foot center of pressure. In addition, individuals responded to medial (lateral) perturbations with a decreased (slightly decreased) hip abduction moment. Contrary to our hypothesis, we did not observe an ankle push-off moment response but rather, a small response in the opposite direction. These results highlight the response of individuals without mobility impairments to unexpected foot-placement perturbations and provide a basis of comparison for those with impaired balance control.

Analysis of the Relative Motion Between the Socket and Residual Limb in Transtibial Amputees While Wearing a Transverse Rotation Adapter.

The coupling between the residual limb and the lower-limb prosthesis is not rigid. As a result, external loading produces movement between the prosthesis and residual limb that can lead to undesirable soft-tissue shear stresses. As these stresses are difficult to measure, limb loading is commonly used as a surrogate. However, the relationship between limb loading and the displacements responsible for those stresses remains unknown. To better understand the limb motion within the socket, an inverse kinematic analysis was performed to estimate the motion between the socket and tibia for 10 individuals with a transtibial amputation performing walking and turning activities at 3 different speeds. The authors estimated the rotational stiffness of the limb-socket body to quantify the limb properties when coupled with the socket and highlight how this approach could help inform prosthetic prescriptions. Results showed that peak transverse displacement had a significant, linear relationship with peak transverse loading. Stiffness of the limb-socket body varied significantly between individuals, activities (walking and turning), and speeds. These results suggest that transverse limb loading can serve as a surrogate for residual-limb shear stress and that the setup of a prosthesis could be individually tailored using standard motion capture and inverse kinematic analyses.

Variables that Influence Basic Prosthetic Mobility in People With Non-Vascular Lower Limb Amputation.

BACKGROUND: There exists a dearth of evidence on rehabilitation factors that influence prosthetic mobility in people with lower limb amputation (LLA). Examining variables that contribute to prosthetic mobility can inform rehabilitation interventions, providing guidance in developing more comprehensive care for these individuals. OBJECTIVE: To determine the influence of modifiable and non-modifiable variables related to LLA and their impact on prosthetic mobility, using the International Classification of Functioning, Disability and Health (ICF) model. Secondarily, to determine if personal factors and self-reported balance and mobility are predictive of Component timed-up-and-go (cTUG) performance. DESIGN: Cross-sectional study of a convenience sample. SETTING: National conference. PARTICIPANTS: People (N=68) with non-vascular causes of unilateral LLA. METHODS: Assessment of anthropometrics, mobility, bilateral hip extensor strength, hip range of motion, single limb balance, and self report measures. Lasso linear regression and extreme gradient boosting analyses were used to determine influence of variables on prosthetic mobility. MAIN OUTCOME MEASURE: Timed performance of the cTUG. RESULTS: The following five variables were found to influence basic prosthetic mobility (P ≤ .05) in people with transtibial amputation: hip extensor strength, hip range of motion, single limb balance, waist circumference, and age. In the transfemoral cohort, number of comorbidities and waist circumference primarily influenced prosthetic mobility. Additionally, 66% of the variance in cTUG total time for the entire sample could be explained by simply regressing on level of amputation, number of comorbidities, age and Activities-specific Balance Confidence scale score, all variables easily collected in a waiting room. CONCLUSION: Variables that are modifiable with physical therapy intervention including hip extensor strength, hip range of motion, single limb balance, and waist circumference significantly influenced basic prosthetic mobility. These variables can be affected by targeted rehabilitation interventions and lifestyle changes. LEVEL OF EVIDENCE: II.

Ideal operating conditions for a variable stiffness transverse plane adapter for individuals with lower-limb amputation.

Transverse plane shear stress between the prosthetic socket and residual limb often results in soft tissue breakdown and discomfort for individuals with lower-limb amputation. To better understand the effects of reduced transverse plane stiffness in the shank of a prosthesis, a second-generation variable stiffness torsion adapter (VSTA II) was tested with individuals with a transtibial amputation (n = 10). Peak transverse plane moments, VSTA II deflection, range of whole body angular momentum (WBAM), ground reaction impulse, joint work, and personal stiffness preference were evaluated at three fixed stiffness levels (compliant: 0.25 Nm/°, intermediate: 0.75 Nm/°, stiff: 1.25 Nm/°) at three walking speeds (self-selected, fast and slow: +/- 20% of self-selected, respectively) while straight-line walking and performing left and right turns. Residual limb loading decreased and VSTA II displacement increased for reductions in stiffness and both metrics increased with increasing walking speed, while ground reaction impulse and joint work were unaffected. The range of WBAM increased with decreased stiffness, which suggests an increased risk of falling when using the VSTA II at lower stiffness settings. Preference testing showed no significant result, but trends for lower stiffness settings when turning and walking at self-selected speeds were noted, as were stiffer settings when walking straight and at faster speeds. These results show that a device with rotational compliance like the VSTA II could reduce loading on the residual limb during straight walking and turning activities and that factors such as walking speed, activity type and user preference can affect the conditions for optimal use.

Stiffness and energy storage characteristics of energy storage and return prosthetic feet.

BACKGROUND: Mechanical properties of prosthetic feet can significantly influence amputee gait, but how they vary with respect to limb loading and orientation is infrequently reported. OBJECTIVE: The objective of this study is to measure stiffness and energy storage characteristics of prosthetic feet across limb loading and a range of orientations experienced in typical gait. STUDY DESIGN: This study included mechanical testing. METHODS: Force-displacement data were collected at combinations of 15 sagittal and 5 coronal orientations and used to calculate stiffness and energy storage across prosthetic feet, stiffness categories, and heel wedge conditions. RESULTS: Stiffness and energy storage were highly non-linear in both the sagittal and coronal planes. Across all prosthetic feet, stiffness decreased with greater heel, forefoot, medial, and lateral orientations, while energy storage increased with forefoot, medial, and lateral loading orientations. Stiffness category was proportional to stiffness and inversely proportional to energy storage. Heel wedge effects were prosthetic foot dependent. CONCLUSION: Orientation, manufacturer, stiffness category, and heel wedge inclusion greatly influenced stiffness and energy storage characteristics. CLINICAL RELEVANCE: These results and an available graphical user interface tool may help improve clinical prescriptions by providing prosthetists with quantitative measures to compare prosthetic feet.

Hip recovery strategy used by below-knee amputees following mediolateral foot perturbations.

Lower-limb amputees have a higher risk of falling compared to non-amputees. Proper regulation of whole-body angular momentum is necessary to prevent falls, particularly in the frontal plane where individuals are most unstable. However, the balance recovery mechanisms used by lower-limb amputees when recovering from a perturbation are not well-understood. This study sought to understand the balance recovery mechanisms used by lower-limb amputees in response to mediolateral foot perturbations by examining changes to frontal plane whole-body angular momentum and hip joint work. These metrics provide a quantitative measure of frontal plane dynamic balance and associated joint contributions required to maintain balance during gait. Nine amputees and 11 non-amputees participated in this study where an unexpected medial or lateral foot placement perturbation occurred immediately prior to heel strike on the residual, sound or non-amputee limbs. Lateral perturbations of all limbs resulted in a reduced range of whole-body angular momentum and increased positive frontal plane hip work in the first half of single limb support. Medial perturbations for all limbs resulted in increased range of whole-body angular momentum and decreased positive frontal plane hip work, also in the first half of single limb support. These results suggest that medial foot placement perturbations are particularly challenging and that hip strategies play an important role in balance recovery. Thus, rehabilitation interventions that focus on hip muscles that regulate mediolateral balance, particularly the hip abductors, and the use of prostheses with active ankle control, may reduce the risk of falls.

Temporal-Spatial Values During a 180° Step Turn in People with Unilateral Lower Limb Amputation.

BACKGROUND: Daily ambulation with a prosthesis often involves turning to negotiate within the home and community environments, however how people with lower limb loss perform turns is infrequently studied. Administering a common clinical outcome measure to capture turn performance data provides a convenient means of assessing this ubiquitous activity. RESEARCH QUESTION: What temporal-spatial parameters are exhibited by people with unilateral lower limb amputation while performing a 180˚ turn task? METHODS: Forty community-ambulating subjects with unilateral lower limb amputation (20 transtibial amputees, 20 transfemoral amputees) performed the Component Timed-Up-and-Go (cTUG) test turning once in each direction, both toward the intact and toward the prosthetic limb. An instrumented walkway captured temporal-spatial parameters during performance of the 180˚ turn task of the cTUG, while a custom iPad application recorded time and number of steps to perform the turn. Comparisons between turn direction and level of amputation during the cTUG and temporal-spatial results were assessed. RESULTS: People with lower limb amputation spent more time on their intact limb while turning than their prosthetic limb regardless of the position of the intact limb, and those with transfemoral amputation spent significantly more time over the intact limb than those with transtibial amputation. Additionally, subjects with transfemoral amputation performed the turn significantly faster when turning with an inner intact limb. SIGNIFICANCE: Amputees use different movement strategies with altered temporal-spatial characteristics to turn depending on the direction of the turn and the level of amputation. Clinical use of the cTUG could provide evidence supporting prosthetic prescription practice and introduction of novel physical therapy interventions for individuals with lower limb amputation.

Identifying classifier input signals to predict a cross-slope during transtibial amputee walking.

Advanced prosthetic foot designs often incorporate mechanisms that adapt to terrain changes in real-time to improve mobility. Early identification of terrain (e.g., cross-slopes) is critical to appropriate adaptation. This study suggests that a simple classifier based on linear discriminant analysis can accurately predict a cross-slope encountered (0°, -15°, 15°) using measurements from the residual limb, primarily from the prosthesis itself. The classifier was trained and tested offline using motion capture and in-pylon sensor data collected during walking trials in mid-swing and early stance. Residual limb kinematics, especially measurements from the foot, shank and ankle, successfully predicted the cross-slope terrain with high accuracy (99%). Although accuracy decreased when predictions were made for test data instead of the training data, the accuracy was still relatively high for one input signal set (>89%) and moderate for three others (>71%). This suggests that classifiers can be designed and generalized to be effective for new conditions and/or subjects. While measurements of shank acceleration and angular velocity from only in-pylon sensors were insufficient to accurately predict the cross-slope terrain, the addition of foot and ankle kinematics from motion capture data allowed accurate terrain prediction. Inversion angular velocity and foot vertical velocity were particularly useful. As in-pylon sensor data and shank kinematics from motion capture appeared interchangeable, combining foot and ankle kinematics from prosthesis-mounted sensors with shank kinematics from in-pylon sensors may provide enough information to accurately predict the terrain.

Foot and Ankle Joint Biomechanical Adaptations to an Unpredictable Coronally Uneven Surface.

Coronally uneven terrain, a common yet challenging feature encountered in daily ambulation, exposes individuals to an increased risk of falling. The foot-ankle complex may adapt to improve balance on uneven terrains, a recovery strategy which may be more challenging in patients with foot-ankle pathologies. A multisegment foot model (MSFM) was used to study the biomechanical adaptations of the foot and ankle joints during a step on a visually obscured, coronally uneven surface. Kinematic, kinetic and in-shoe pressure data were collected as ten participants walked on an instrumented walkway with a surface randomly positioned ±15 deg or 0 deg in the coronal plane. Coronally uneven surfaces altered hindfoot-tibia loading, with more conformation to the surface in early than late stance. Distinct loading changes occurred for the forefoot-hindfoot joint in early and late stance, despite smaller surface conformations. Hindfoot-tibia power at opposite heel contact (@OHC) was generated and increased on both uneven surfaces, whereas forefoot-hindfoot power was absorbed and remained consistent across surfaces. Push-off work increased for the hindfoot-tibia joint on the everted surface and for the forefoot-hindfoot joint on the inverted surface. Net work across joints was generated for both uneven surfaces, while absorbed on flat terrain. The partial decoupling and joint-specific biomechanical adaptations on uneven surfaces suggest that multi-articulating interventions such as prosthetic devices and arthroplasty may improve ambulation for mobility-impaired individuals on coronally uneven terrain.

Turn Intent Detection For Control of a Lower Limb Prosthesis.

OBJECTIVE: An adaptable lower limb prosthesis with variable stiffness in the transverse plane requires a control method to effect changes in real time during amputee turning. This study aimed to identify classification algorithms that can accurately predict turning using inertial measurement unit (IMU) signals from the shank with adequate time to enact a change in stiffness during the swing phase of gait when the prosthesis is unloaded. METHODS: To identify if a turning step is imminent, classification models were developed around activities of daily living including 90° spin turns, 90° step turns, 180° turns, and straight walking using simulated IMU data from the prosthesis shank. Three classifiers were tested: support vector machine (SVM), K nearest neighbors (KNN), and a bagged decision tree ensemble (Ensemble). RESULTS: Individual training gave superior results over training on a pooled set of users. Coupled with a simple control scheme, the SVM, KNN, and Ensemble classifiers achieved 96%, 93%, and 91% accuracy (no significant difference), respectively, predicting an upcoming turn 400 ± 70 ms prior to the heel strike of the turn. However, classification of straight walking transition steps varied between classifiers at 85%, 82%, 97% (Ensemble significantly different, ), respectively. CONCLUSION: The Ensemble model produced the best result overall; however, depending on the priority of identifying turning versus transition steps and processor performance, the SVM or KNN might still be considered. SIGNIFICANCE: This research would be useful to help determine a classifier strategy for any lower limb device seeking to predict turn intent.

The effects of prosthetic foot stiffness on transtibial amputee walking mechanics and balance control during turning.

BACKGROUND: Little evidence exists regarding how prosthesis design characteristics affect performance in tasks that challenge mediolateral balance such as turning. This study assesses the influence of prosthetic foot stiffness on amputee walking mechanics and balance control during a continuous turning task. METHODS: Three-dimensional kinematic and kinetic data were collected from eight unilateral transtibial amputees as they walked overground at self-selected speed clockwise and counterclockwise around a 1-meter circle and along a straight line. Subjects performed the walking tasks wearing three different ankle-foot prostheses that spanned a range of sagittal- and coronal-plane stiffness levels. FINDINGS: A decrease in stiffness increased residual ankle dorsiflexion (10-13°), caused smaller adaptations (<5°) in proximal joint angles, decreased residual and increased intact limb body support, increased residual limb propulsion and increased intact limb braking for all tasks. While changes in sagittal-plane joint work due to decreased stiffness were generally consistent across tasks, effects on coronal-plane hip work were task-dependent. When the residual limb was on the inside of the turn and during straight-line walking, coronal-plane hip work increased and coronal-plane peak-to-peak range of whole-body angular momentum decreased with decreased stiffness. INTERPRETATION: Changes in sagittal-plane kinematics and kinetics were similar to those previously observed in straight-line walking. Mediolateral balance improved with decreased stiffness, but adaptations in coronal-plane angles, work and ground reaction force impulses were less systematic than those in sagittal-plane measures. Effects of stiffness varied with the residual limb inside versus outside the turn, which suggests that actively adjusting stiffness to turn direction may be beneficial.

Second generation prototype of a variable stiffness transverse plane adapter for a lower limb prosthesis.

The prescription and fitting of a traditional lower limb prosthesis often focuses on straight walking, neglecting maneuvers such as turning and pivoting that require flexibility in the transverse plane. Current devices that allow transverse plane movement only offer a fixed stiffness and are incapable of adapting to varying daily activities. Pilot testing of a first-generation variable stiffness torsion adapter (VSTA I) showed a benefit for individuals with lower limb amputation by reducing peak transverse plane moments at the residual limb which could lead to increased comfort, but testing was limited due to excessive device height and mass. The VSTA II, a second-generation prototype, is capable of discrete stiffness variation from 0.31 to 1.29 Nm/° in 0.25 Nm/° increments with ±30° of motion in addition to fully locked operation. Stiffness variation is enabled by five independent spring subunits that can be combined in parallel to create different, linear, stiffness settings. The VSTA II features a reduced mass (51% reduction) and height (42% reduction) compared to its predecessor along with a tether-free controller and power system. These improvements will permit greater recruitment for amputee studies, and allow for advanced testing both in and out of the lab.