CARE-PD is the largest publicly available 3D gait mesh dataset for Parkinson's disease, released in 2025 by the University of Toronto in collaboration with the Vector Institute, KITE Research Institute–UHN, and other institutions. The related research paper is titled "CARE-PD: A Multi-Site Anonymized Clinical Dataset for Parkinson's Disease Gait AssessmentThe data has been selected for NeurIPS 2025 Datasets and Benchmarks, aiming to provide a high-quality data foundation for clinical score prediction, Parkinson's gait representation learning, and unified analysis across institutions.

This dataset contains gait records from 362 subjects across 9 independent cohorts at 8 clinical institutions. All gait videos and motion capture data were uniformly processed and converted into anonymized SMPL 3D human gait meshes.

Dataset composition

This dataset consists of nine uniformly processed subsets, each containing different clinical information or motion features:

• 3DGait: Clinical gait recording, including UPDRS score
• BMCLab: Includes drug status and UPDRS scores (Original license: CC BY 4.0)
• DNE: Includes healthy subjects, PD patients, and other neurological disorders (Original license: CC BY 4.0)
• E-LC: Includes drug status (on/off) and PD subtype
• KUL-DT-T: Includes frozen/non-frozen gait subtypes
• PD-GaM: Clinical gait record, including UPDRS score
• T-SDU: Gait Recording in Everyday Environments
• T-SDU-PD: Gait in PD patients, including UPDRS score.
• T-LTC: Gait Recording in Everyday Environments