MotionInsight - Database

Motion Insight Hub: Dataset for Dyskinesia Assessment in Parkinson’s Disease This dataset contains anonymized upper‑limb motion data collected for the study “A Technical Framework for Detecting Dyskinesia in Parkinson’s Disease Using Motion Tracking and Serious Games” (PeerJ, submitted). Data were acquired using Motion Insight Hub, a research web platform developed by the SA‑BIO Research Group at Universidad Rey Juan Carlos (URJC), Madrid, Spain. The official project website is: https://motioninsighthub.web.app/ The study integrates custom-designed serious games with contactless motion tracking based on the Leap Motion Controller (v1.0) to support quantitative assessment of Parkinsonian motor dysfunction, with a particular focus on dyskinesia, bradykinesia, tremor, coordination, and fatigue‑induced motor instability. The dataset includes recordings from 46 participants (20 individuals diagnosed with Parkinson’s disease and 26 normative participants). For machine‑learning analyses, a subset of 44 participants was used, excluding two young reference users included only for usability evaluation. Parkinson’s disease severity spans Hoehn and Yahr stages 1.0 to 3.0. Participants interacted with three exergames designed to elicit distinct motor patterns: A baseline static task to assess resting tremor and involuntary movements. A bilateral hand open/close task to quantify bradykinesia, dexterity, and symmetry. A vertical obstacle‑avoidance task to evaluate reaction time, coordination, and fatigue‑related dyskinesia, implemented in three visual variants. Raw data consist of frame‑level CSV recordings captured without physical contact, including palm position, hand speed, orientation vectors, fingertip coordinates, grip strength, timestamps, and device metadata. From these recordings, 62 kinematic, temporal, and spectral features were derived and used for statistical analysis and machine‑learning classification. All data are fully anonymized. No personally identifiable information, video recordings, or images are included. The protocol was approved by the Ethics Committee of Universidad Politécnica de Madrid (UPM), and all participants provided informed consent. This dataset is intended to support reproducibility, methodological comparison, and research on digital biomarkers, serious games, and human–computer interaction in Parkinson’s disease. It is released exclusively for non‑commercial research and educational use and is not intended for clinical diagnosis or medical decision‑making. Project website: The official project website is: 👉 MotionInsight - Database The website provides:- interactive visualisations of the recorded motion data,- exploratory plots and summaries of kinematic variables,- public dissemination material related to the project. Funding:This work was supported by the State Research Agency of Spain (Agencia Estatal de Investigación, AEI) under Grant CarHaVoz: PID2023‑152984OB‑I00. Contact:SA‑BIO Research Group, Universidad Rey Juan Carlos (URJC), Madrid, SpainCorresponding author: Daniel Palacios‑Alonso ([e-mail])