| Literature DB >> 33742069 |
Solveig K Sieberts1, Jennifer Schaff2, Marlena Duda3, Bálint Ármin Pataki4, Ming Sun5, Phil Snyder6, Jean-Francois Daneault7,8, Federico Parisi7,9, Gianluca Costante7,9, Udi Rubin10, Peter Banda11, Yooree Chae6, Elias Chaibub Neto6, E Ray Dorsey12, Zafer Aydın13, Aipeng Chen14, Laura L Elo15, Carlos Espino10, Enrico Glaab11, Ethan Goan16, Fatemeh Noushin Golabchi7, Yasin Görmez13, Maria K Jaakkola15,17, Jitendra Jonnagaddala18,19, Riku Klén15, Dongmei Li20, Christian McDaniel21,22, Dimitri Perrin23, Thanneer M Perumal6, Nastaran Mohammadian Rad24,25,26, Erin Rainaldi27, Stefano Sapienza7, Patrick Schwab28, Nikolai Shokhirev10, Mikko S Venäläinen15, Gloria Vergara-Diaz7, Yuqian Zhang29, Yuanjia Wang30, Yuanfang Guan3, Daniela Brunner10,31, Paolo Bonato7,9, Lara M Mangravite6, Larsson Omberg32.
Abstract
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).Entities:
Year: 2021 PMID： 33742069 PMCID： PMC7979931 DOI： 10.1038/s41746-021-00414-7
Source DB: PubMed Journal: NPJ Digit Med ISSN： 2398-6352