Interpretable Side-Aware Gait-State Representations for Post-Stroke Neurorobotics

Rocco Salvatore Calabrò published a secondary analysis of a public multimodal gait dataset on May 25, 2026. The study used data from 138 able-bodied adults and 50 adults with stroke to develop side-aware gait-state representations. The analysis focused on 11 waveform domains, including sagittal kinematics and surface electromyography, to identify three distinct gait states in a cohort of 43 stroke patients.

Use Cases

• Developing transparent gait-state representations for adaptive lower-limb neurorobotics based on multimodal waveform data.
• Analyzing side-dependent differences in post-stroke gait using paretic and non-paretic waveform views.
• Clustering post-stroke walking patterns into interpretable states based on kinematic and sEMG data.
• Hypothesis generation for future clinical studies on neurorobotic state-representation design.

Strengths

• Analysis is based on a public dataset with 188 total participants (138 able-bodied, 50 with stroke).
• The retained fused solution organizes a strict complete-case stroke cohort of 43 individuals into three interpretable states.
• The study identifies key contributing domains, including ankle angle, vastus lateralis sEMG, and knee angle.

Limitations

• The dataset is very small at 30.8 KB, suggesting it contains analysis documentation rather than raw data.
• Column-level documentation is absent; field semantics must be inferred after download.
• The description indicates the contribution is exploratory and representational, not validated for clinical or real-time use.

Provenance

Source

Rocco Salvatore Calabrò via figshare

Collection Method

Secondary analysis of a public multimodal gait dataset.

Freshness

Last updated 2026-05-25 06:05:42