Tools And Standards For Digital Biomarker Discovery
MENU
FOMO: Functions for Optimizing Missing Observations
We study how missing data in Bring-Your-Own-Device (BYOD) wearable studies relates to health learnings. In many wearable studies, gaps in data are treated as a technical problem to be ignored or patched over. However, there is no clear standard for handling missing data, and these gaps can quietly distort study conclusions. In this project, we examine when and how wearable data go missing, from short gaps during the day to people stopping use altogether, and identify common patterns in device adherence. We then assess how these patterns can bias study conclusions and link missingness to clinical characteristics. Our goal is to establish practical guidelines for wearables data quality, determine whether “nuisance” signals like missingness may actually be informative, and improve the reliability of health insights drawn from real-world wearables studies.
Check out our recent workshop from IEEE-EMBS BHI 2025 in Atlanta, Georgia!
Related Publications: Dunn, Jessilyn, Leeor Hershkovich, Hayoung Jeong, Shun Sakai, Harrison Kane, Andrew Allen, and Benjamin Goldstein. “Defining the Habitome: Phenotypes of Routine and Their Relationship to Health Outcomes.” Research Square, 2025. https://doi.org/10.21203/rs.3.rs-5861743/v2
Lab members: Hayoung Jeong, Leeor Hershkovich, Harrison Kane
DBDP: Digital Biomarker Discovery Project
The Digital Biomarker Discovery Project (DBDP) aims to improve the transparency and reproducibility of digital biomarker studies by providing open-source frameworks and analytical methods. We also provide educational materials for both new and seasoned digital health researchers. There are ample opportunities to collaborate with the DBDP team across data science, biomedical engineering, and clinical research while contributing to a growing ecosystem for digital health innovation.
Lab members: Bill Chen, Hayoung Jeong, Anita Silver Shlesinger
JupyterHealth
The BIG IDEAs lab is in collaboration with experts from Berkeley, UCSF, and Project Jupyter to build Jupyter Health. This innovative open-source platform aims to enhance the secure collection of data from wearable sensors to electronic health record and to streamline the development, sharing, and validation of digital biomarker algorithms. We develop modular features for Jupyter Health, ranging from comprehensive summary statistics to dynamic user visualizations that can provide immediate insights into health data. While we are currently piloting Jupyter Health development for diabetes management, we envision Jupyter Health one day covering a wide array of health conditions.
Lab members: Bill Chen
