LegPrio study data

# Proprioception Assessment – Vicon Motion Capture Dataset This repository contains the dataset used in the final degree project: **"Wearable System for Proprioception Assessment and Step Width Inference"** by Martí Mestre, Universitat Politècnica de Catalunya (UPC), 2025. The dataset includes `.mat` files exported from the **Vicon Motion Capture System**, capturing 5 trials per subject under controlled walking conditions. The data was used to analyze proprioceptive errors, validate a wearable sensor system, and simulate step width prediction models. --- ## 📁 File Naming Convention Each file follows this format:S<subject_number>Crisscross_T<trial_number>.mat | Segment | Meaning ||-----------------|----------------------------------------------|| `S3013` | Subject ID || `T1` to `T5` | Trial number || `right`, `left`, etc. | Walking condition (see below) | ### 🔍 Example Filenames - `S3013_Crisscross_T1_right.mat` — Right leg crossing over left at 0.7 m/s - `S3013_Crisscross_T2_left.mat` — Left leg crossing over right at 0.7 m/s - `S3013_Crisscross_T3_right.mat` — Right over left at 0.9 m/s - `S3013_Crisscross_T4_right.mat` — Right over left at 0.5 m/s - `S3013_Crisscross_T5_right.mat` — Right over left at 0.7 m/s, eyes closed > All trials are 2 minutes long, sampled at 150 Hz (~18,000 frames). --- ## 🧪 Trial Descriptions | Trial | Condition | Description ||-------|------------------------|----------------------------------------------|| T1 | `right` | 0.7 m/s — Right leg crossing over left || T2 | `left` | 0.7 m/s — Left leg crossing over right || T3 | `right_09ms` | 0.9 m/s — Right leg crossing over left || T4 | `right_05ms` | 0.5 m/s — Right leg crossing over left || T5 | `right_closed` | 0.7 m/s — Right over left with eyes closed | --- ## 📦 File Contents Each `.mat` file contains a structured Vicon export with the following top-level fields: | Field | Description ||--------------|-----------------------------------------------------------------------------|| `Model` | Includes marker set (`MkrSet`) and subject ID (`Subject`) || `Analog` | Force plate signals (e.g., Bertec), units, labels || `FPlate` | Force plate geometry, origin, and type (e.g., AMTI or Bertec) || `Coor` | 3D marker trajectories (`Signal`: [frames × markers × 3]) || `Processing` | Biomechanical outputs including joint angles, gait events, and anthropometry | ### 🧠 Example `Processing` Fields - `Bodymass`, `Height`- `RSHN_R_Tibia_Z`, `LTHI_L_Femur_X`, etc. (segment angles)- Heel strike and toe-off event timing (if available) --- ## 🔬 Usage This dataset was used to:- Evaluate proprioceptive error in leg-crossing detection- Validate a 3D-printed wearable device with 9 distance sensors and 1 IMU- Train and test polynomial models to infer ankle distance across the gait cycle- Justify the need for step-width inference instead of direct measurement --- ## 📄 Citation If you use this dataset, please cite:Martí Mestre, Maria Artigas Trayter, Final Degree Project: "Wearable System for Proprioception Assessment and Step Width Inference," Universitat Politècnica de Catalunya, 2025. --- ## 📬 Contact **Martí Mestre** 📧 [e-mail] 🔗 GitHub: [https://github.com/martimestre](https://github.com/martimestre) **Maria Artigas Trayter** 📧 [e-mail]