Answers are grounded in this paper and cite their sources.
This study evaluated personalized gait phase estimation for stroke survivors using functional inertial measurement unit (IMU) alignment and two-stage supervised personalization of models pre-trained on healthy gait. The estimator used a thigh-mounted IMU, and heel forcesensitive resistor measurements provided reference phase labels for offline adaptation and evaluation. Stage 1 established a distillation-regularized participant-specific model, and Stage 2 performed conditional refinement using low-rank adaptation. Long Short-Term Memory (LSTM), Temporal Convolutional Network (TCN), and Transformer models were evaluated in five stroke survivors walking with a powered knee exoskeleton using leave-one-subject-out hyperparameter selection and sequential test-then-adapt replay. Relative to non-adapted baselines, Stage 1+2 reduced the mean participant-wise phase root mean square error by 84.2%, 77.0%, and 60.7%, respectively. The Transformer achieved the lowest final error (2.90 ± 1.13% of the gait cycle) and heel-strike timing error (23.7 ± 4.5 ms). Ablations showed no group-mean accuracy advantage of distillation regularization over standard fine-tuning. Every-cycle updating generally minimized error, whereas conditional updating reduced update counts with small accuracy differences, demonstrating an accuracy-update-frequency trade-off. After personalization, alignment produced model-dependent changes in phase error while preserving or improving heel-strike detection and reducing heel-strike timing error for the LSTM and Transformer. In concurrent embedded tests, the TCN and Transformer maintained 100-Hz inference during Stage 2 updates without deadline misses, whereas the LSTM missed the 10-ms deadline in 6.6% of inferences. All updates completed within 0.8 s. These results support the offline feasibility and embedded timing of the two-stage framework for exoskeleton-assisted walking.
No takes yet. Share an insight, caveat, or question.
Ryu et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: