Aydin, ZaferSieberts, Solveig K.Schaff, JenniferDuda, MarlenaPataki, Balint ArminSun, MingSnyder, PhilDaneault, Jean-FrancoisParisi, FedericoCostante, GianlucaRubin, UdiBanda, PeterChae, YooreeChaibub Neto, EliasDorsey, E. RayChen, AipengElo, Laura L.Espino, CarlosGlaab, EnricoGoan, EthanGolabchi, Fatemeh NoushinGormez, YasinJaakkola, Maria K.Jonnagaddala, JitendraKlen, RikuLi, DongmeiMcDaniel, ChristianPerrin, DimitriPerumal, Thanneer M.Rad, Nastaran MohammadianRainaldi, ErinSapienza, StefanoSchwab, PatrickShokhirev, NikolaiVenalainen, Mikko S.Vergara-Diaz, GloriaZhang, YuqianWang, YuanjiaGuan, YuanfangBrunner, DanielaBonato, PaoloMangravite, Lara M.Omberg, Larsson2022-03-052022-03-0520212398-6352PubMed ID33742069https //doi.org/10.1038/s41746-021-00414-7https://hdl.handle.net/20.500.12573/1245The Parkinson's Disease Digital Biomarker Challenge was funded by the Robert Wood Johnson Foundation and the Michael J. Fox Foundation. Data were contributed by users of the Parkinson mPower mobile application as part of the mPower study developed by Sage Bionetworks and described in Synapse [https://doi.org/10.7303/syn4993293].Resources and support for J.S. were provided by Elder Research, an AI and Data Science consulting agency. M.D. was supported by NIH NIGMS Bioinformatics Training Grant (5T32GM070449-12). J.F.D. was supported by a postdoctoral fellowship from the Canadian Institutes of Health Research. L.L.E. reports grants from the European Research Council ERC (677943), European Union's Horizon 2020 research and innovation programme (675395), Academy of Finland (296801, 304995, 310561 and 314443), and Sigrid Juselius Foundation, during the conduct of the study. EG1 acknowledges the funding support by the Fonds Nationale de la Recherche (FNR) Luxembourg, through the National Centre of Excellence in Research (NCER) on Parkinson's disease (I1R-BIC-PFN-15NCER), and as part of the grant project PD-Strat (INTER/11651464). M.K.J. was supported by Alfred Kordelin Foundation. J.J. is supported by UNSW Sydney Electronic Practice Based Research Network (ePBRN) and Translational Cancer Research Network (TCRN) programs. D.L. is supported in part by the University of Rochester CTSA award number UL1 TR002001 from the National Center for Advancing Translational Sciences of the National Institutes of Health. PS2 is supported by the Swiss National Science Foundation (SNSF) project No. 167302 within the National Research Program (NRP) 75 "Big Data". P.S. is an affiliated PhD fellow at the Max Planck ETH Center for Learning Systems. Y.G. is supported by NIH R35GM133346, NSF#1452656, Michael J. Fox Foundation #17373, American Parkinson Disease Association AWD007950. Cohen Veterans Bioscience contributed financial support to Early Signal Foundation's costs (U.R., C.E., and D.B.).Consumer wearables and sensors are a rich source of data about patients' daily disease and symptom burden, particularly in the case of movement disorders like Parkinson's disease (PD). However, interpreting these complex data into so-called digital biomarkers requires complicated analytical approaches, and validating these biomarkers requires sufficient data and unbiased evaluation methods. Here we describe the use of crowdsourcing to specifically evaluate and benchmark features derived from accelerometer and gyroscope data in two different datasets to predict the presence of PD and severity of three PD symptoms: tremor, dyskinesia, and bradykinesia. Forty teams from around the world submitted features, and achieved drastically improved predictive performance for PD status (best AUROC = 0.87), as well as tremor- (best AUPR = 0.75), dyskinesia- (best AUPR = 0.48) and bradykinesia-severity (best AUPR = 0.95).enginfo:eu-repo/semantics/openAccessGENDER-DIFFERENCESHYPOTHESIS TESTSVALIDATIONarticleVolume 4 Issue 1