Fig 1.
Musculoskeletal models with bilateral AFOs.
The beginning of second double-limb support is shown for one typically-developing (TD) participant and one participant from each level of crouch severity. Mild (MI), moderate (MO) and severe (SE) crouch gait were defined by the minimum knee flexion angle during stance. AFO torque (τAFO) was determined by AFO stiffness and AFO angle (θAFO) for the passive AFOs, and by the OpenSim cost function for the powered AFOs.
Table 1.
Participants (average ± one standard deviation).
Fig 2.
Sagittal-plane joint kinematics and internal moments.
Top: Ankle, knee and hip kinematics for gait in TD children and children with crouch gait. TD children walked with less ankle dorsiflexion and knee flexion during stance than those with crouch gait. Bottom: Ankle, knee and hip moments for gait in TD children and crouch gait. TD children generated larger peak plantarflexor moments and smaller peak knee extensor moments compared to crouch gait. Knee extensor moments increased with increasing crouch severity.
Fig 3.
Simulation pipeline and outcome measures.
Leg (Lleg,j,k) and muscle (Li,k, Li,j) impulses were computed for all AFO conditions, including the baseline (Unassisted) condition. Subscripts i, j, k, denote the muscle, AFO condition, and mass, respectively. The muscle impulses used to compare between AFO conditions were computed from the beginning of single-limb support (tSS) to the start of swing (tSW). Muscle impulses for the mass analysis (black dotted line) were computed from tSW to initial contact at the start of the next gait cycle (tIC). Forces were normalized by bodyweight (BW). Leg impulse was computed for each participant’s leg that contained the single-limb support and second double-limb support gait phases. Change (ΔLi,j,k) and percent change (%ΔLi,j,k) in leg and muscle impulses were used to quantify changes in muscle demand between AFO conditions. Abbreviations: Lleg,i,j,k, Leg impulse for muscle i, AFO j, and mass k; SS, single-limb support phase; SW, swing phase; IC, initial contact; Δ, absolute change; %Δ, percent change.
Fig 4.
Leg muscle force with each AFO condition compared to unassisted walking.
Profiles are averaged across participants in each group. Top to bottom: Optimal passive AFO, unidirectional powered AFO, and bidirectional powered AFO. The integral of these curves represents the leg impulse. Reductions in leg muscle force occurred primarily during late stance (40–60% gait cycle) for the TD group, while crouch gait groups saw leg muscle force reduced throughout single-limb support and late stance (20–60% gait cycle). Passive AFOs reduced leg muscle force less than unidirectional powered AFOs during early single-limb support for all groups, and throughout stance for the TD and moderate crouch groups. For some participants, passive and unidirectional powered AFO torque profiles were nearly identical, resulting in only small differences in leg muscle force with different AFOs. Reductions in leg muscle force were nearly identical for the unidirectional and bidirectional powered AFO conditions. Small differences in leg muscle force between these conditions occurred during swing and corresponded to changes in tibialis anterior force due to dorsiflexion assistance in the bidirectional AFO.
Fig 5.
AFO torque profiles, leg impulse for each AFO condition and predictors of reductions in leg and muscle impulses.
Top: Net ankle moments determined by inverse dynamics and AFO torque profiles for gait in TD children and children with crouch gait. A positive moment corresponds to a plantarflexor torque. Bottom, left: Leg impulse magnitude increased with crouch severity. Bottom, center: Reduction in leg impulse was strongly correlated with nondimensional speed for all AFO conditions. Bottom, right: GAS impulse was most strongly correlated with peak knee flexor moment. GAS activity in one subject with severe crouch was estimated to be near-zero during stance, and was omitted from this figure. Abbreviations: TD, typically-developing; MI, mild crouch; MO moderate crouch; SE; severe crouch; GAS, gastrocnemius muscle group.
Table 2.
Reduction in leg impulse versus unassisted gait, showing average absolute (xBW, ± SD) and percent change (± SD) in leg impulse.
Fig 6.
Percent change in impulse of individual muscles across AFO conditions.
The GAS and SOL were most impacted by all AFO conditions, with relatively small changes occurring in other muscles. Top: The TA impulse increased with passive AFOs to overcome the AFO’s dorsiflexion resistance and maintain ankle kinematics. The large percent increase in TA impulse for the optimized passive AFO corresponds to only a small absolute increase in TA impulse compared to unassisted gait. Middle: The unidirectional powered AFO had similar reductions in muscle impulses as the bidirectional AFO, except for the GAS. Bottom: Only the bidirectional powered AFO reduced TA impulse, but this corresponded to a smaller percent reduction in GAS impulse. Impulses in muscles spanning the knee and hip changed by less than 20%, with the VAS having the largest reductions in muscle demand within these groups. Abbreviations: GAS, gastrocnemius; SOL, soleus; TA, tibialis anterior; VAS, vasti; RF, rectus femoris; HAMS, biarticular hamstrings; GMAX, gluteus maximus; ILIO, iliopsoas.
Table 3.
Sensitivity of leg impulse to AFO mass (xBW/kgAFO).