(K-417) Effects of Assistive and Resistive Torques at the Knee during Overground Walking in Children and Young Adults: A multi-visit study

Introduction:: Cerebral palsy (CP) and spina bifida (SB) are prevalent causes of physical disability in children. Both can cause crouch gait, a pathological walking pattern characterized primarily by increased flexion of the knee as well as the ankle and hip, which is less functional, more energetically taxing, and can lead to joint pain and degradation. Although not neurologically progressive diseases, walking ability often worsens with age and many lose the ability to walk into adulthood. Thus, intensive gait training in childhood is critical to retaining mobility.

Current standard of care has mixed results; thus, a need exists for new options. Exoskeletons are becoming increasingly prominent in rehabilitation and present a promising alternative to existing procedures. Exoskeletons can be used as either assistive or rehabilitation devices. The former approach can make walking easier by applying torque to straighten the legs to promote stance or bend the legs to step. For rehabilitation, the exoskeleton can provide resistance to make walking harder, with the goal of increasing muscle activation and strength over time. These two approaches can be combined whereby assistance is provided during more challenging phases of the gait cycle while resistance is applied at other points. This strategy may be particularly suitable for children who are more impaired and are otherwise unable to walk under the presence of resistance. The goal of this work was to investigate the effects of assistive and resistive torques applied at the knee in children and young adults with crouch gait across multiple visits using the exoskeleton.

Materials and Methods:: We performed an observational study wherein participants with crouch from CP/SB practice walking with our novel exoskeleton which comprises a custom fit knee-ankle-foot orthosis (KAFO), an actuator (Agilik®, Bionic Power) at the knee, and a passive ankle joint.

The exoskeleton measures knee angle and angular velocity, and foot-ground contact in real-time to split the gait cycle into 5 phases: early/mid/late stance, and early/late swing. Within each phase, assistive or resistive torque is applied to the knee. Three modes were evaluated. The zero mode is the baseline with net zero knee torque to compensate for exoskeleton inertia. In assist mode, extension torque is provided during early and mid-stance and late swing. In interleaved mode, knee extension assistance is applied during mid-stance and resistance is applied during late swing.

The study is 10 total visits. The first two are for exoskeleton fitting and setup. Remaining visits include walking practice, with data collection at the 3rd, 7th and 10th visits including motion capture and electromyography (EMG) recorded at 100 and 2000 Hz, respectively.

Data from the first and last assessments from 3 participants with CP (ages: 7, 13, and 25 years) and 1 with SB (9 years) (all GMFCS III) are analyzed in this interim analysis. All participants gave informed consent. Outcomes are peak knee extension during stance, knee range of motion (ROM), and vastus lateralis activation during swing assessed as area under the EMG curve. For each visit, EMG was normalized by zero mode EMG for comparison across visits and participants.

Results, Conclusions, and Discussions::

As in our prior studies, the exoskeleton had an immediate effect on peak knee extension in the more affected (MA) leg with both the interleaved (27.3°±15.8°) and assist (19.5°±15.7°) modes showing increased knee extension compared to the zero mode (28.7°±11.4) (Fig. 1). Knee ROM also improved (Fig. 2). Comparing the first and last assessments, separated by 6 practice sessions, the largest improvements were observed in the interleaved mode with peak knee extension angle improving on average by 7.8°±18.7° and 10.7°±14.7° (Fig. 1) for the MA and less affected (LA) leg respectively, whereas knee ROM increased by 13.4°±12.1° and 13.1°±11.5° (Fig. 2) for the MA and LA leg, respectively. The zero mode also showed improvement from the first to last visit in peak knee extension (MA: 4.9°±12.5° LA: 2.2°±14.4°) (Fig. 1) and knee ROM (MA: 7.6°±10.5° LA: 9.6°±9.1°) (Fig. 2) for both legs. Initially, vastus lateralis (VL) activity during swing with assist and interleaved modes was greater than zero mode, but at the final visit VL activity was lower in the assist compared to zero mode suggesting the users’ improved limb posture resulted in reduced muscle activity after acclimating to the exoskeleton. VL activity with the interleaved mode was still higher than zero mode at the final assessment (Fig. 3), indicating the exoskeleton can facilitate exercise, even for participants who would be otherwise unable to handle such resistance.

Collectively, these results indicate that the assist and interleaved modes improve knee extension and knee ROM during walking in individuals with crouch gait. Further, our initial results here from this ongoing study show knee extension is improved between first and last visits in the zero mode, suggesting that the benefits of the 9 practice sessions may carry over without the exoskeleton. Therefore, longer term practice with the exoskeleton may further improve walking. Importantly, the response to each exoskeleton mode varied across participants, and even between limbs within each individual, which was expected due to the heterogeneity of CP and SB. Thus, tailoring the operational mode to each user is important for maximizing the functional and potential therapeutic benefits of exoskeleton mediated gait training.

Acknowledgements (Optional): :

References (Optional): :