Title: Multimodal Synchronized Motion Capture, Force Plate, and Radar Dataset of the One-Legged Stand Test for Fall Risk Assessment
Description:
This dataset contains synchronized recordings of 32 healthy participants (15 young adults ≤32y, 17 older adults ≥64y) performing the One-Legged Stand Test (OLST), a validated clinical assessment used to evaluate fall risk and postural control. The recordings include three sensor modalities:
1. Motion capture (MOCAP) at 100 Hz using a Qualisys marker-based system.
2. Force plates at 1200 Hz from dual plates under each foot.
3. 24 GHz Frequency-Modulated Continuous Wave (FMCW) radar data at ~27.6 Hz.
Each OLST attempt is labeled with four key postural events:
- Foot-lift
- Start-of-stability
- End-of-stability
- Foot-touchdown
These annotations are provided in both MOCAP/force-plate timestamps and radar frame indices. The older adult participants completed longer 20-second trials to assess sustained balance. The dataset is suitable for research in biomechanics, balance analysis, radar signal processing, and digital biomarker development.
Directory Structure:
- `Metadata/`
Contains demographic information, sensor settings, marker configurations, and event-level metadata (`OLST_Attempts.csv`).
- `OLST_QTM/`
Contains the full Qualisys Track Manager (QTM) project and synchronized 3D motion capture recordings in `.qtm` format.
- `Raw/`
Includes unprocessed MOCAP (`/MOCAP/`) and force plate (`/ForcePlate/`) recordings in `.csv` format, organized by participant ID.
- `Processed/Radar_RDMs/`
Contains processed radar Range Doppler Maps (`.mat` files), cropped and annotated for each capture.
- `Code/`
Contains a multi-modal data loading and visualization Python script (viz_OLST_attempt.ipynb) as an example of working with the dataset.
File Naming Convention:
Files follow the structure:
`ParticipantID_MovementCode_Modality_VersionNumber`
Example:
'01_MNTRR_RR_V2': Participant 01; Mountain-to-Tree (right foot as base); radar recording; version 2.
Utility Notebook:
In viz_OLST_attempt.ipynb, basic Python functions are included for:
- Visualizing OLST attempts
- Accessing and aligning multimodal data
- Converting radar frames to time using `Seconds_per_Frame`
Getting Started:
1. Download the full dataset and unzip all folders.
2. Open the QTM project using `Settings.qtmproj` in the `OLST_QTM/` folder to visualize motion capture and force data. (optional)
3. Use the `viz_OLST_attempt.ipynb` Jupyter notebook to load metadata and event labels from `OLST_Attempts.csv`. This generates a synchronized visualization of MOCAP, force plate, and radar data for a single OLST attempt.
Notes:
- Capture `12_MNTRL_V2` is missing due to a data collection error.
- Radar frame timing varies slightly across captures; use the provided `Seconds_per_Frame` field for accurate time alignment.
- Only long-duration (20 s) OLST trials are available for older adults.
Ethics:
The study was approved by the MIT Committee on the Use of Humans as Experimental Subjects (COUHES #1911000055), and all participants provided written informed consent. All data have been de-identified in accordance with HIPAA Safe Harbor guidelines.
Citation:
If you use this dataset, please cite the associated data descriptor paper:
Copeland D., Zhang X., Linton E., Mori B., Lugaro H., Anthony B.W.
"Multimodal Synchronized Motion Capture, Force Plate, and Radar Dataset of the One-Legged Stand Test for Fall Risk Assessment." *Scientific Data*, 2025. Under review.
Contact:
Daniel Copeland – dcope@mit.edu
MIT Center for Clinical and Translational Research