Data Science and Analytics for Precision Rehabilitation (DAPR) Center - Resource Core

University Of Southern California, Los Angeles CA

Investigators

Abstract

PROJECT SUMMARY – RESOURCE CORE The Data Science and Analytics for Precision Rehabilitation (DAPR) Center’s Resource Core will encourage the uptake of DAPR tools and promote general data science, AI, and ML literacy in the rehabilitation community by providing innovative, accessible resources. Building on the success of ReproRehab (NIH-NCMRR R25HD105583), the core will provide hands-on services, such as drop-in office hours and consultations on data science, AI, and ML, and educational content like software, code, tutorials, and living guidebooks. The Resource Core will collaborate with the Community Engagement and Outreach and Administrative Cores to manage and disseminate resources to the rehabilitation research community via online and hybrid methods. The Resource Core will accomplish two specific aims. Aim 1 is to provide hands-on services to support rehabilitation-specific data harmonization, data sharing, and precision rehabilitation algorithm development. Free one-on-one, extensive project consultations and drop-in office hours will be offered to the broader rehabilitation research community, helping rehabilitation community members to develop DAPR data schemas for new conditions, develop new specifications for rehabilitation data elements that do not yet exist, and implement new software features in Nipoppy and Neurobagel to annotate, manage, and query datasets with the DAPR data schema. Consultations and office hours will also provide general assistance with organizing, managing, and archiving rehabilitation datasets, developing precision rehabilitation algorithms, and sharing data and code. Dedicated support will be provided to pilot grant awardees to ensure project success, facilitating knowledge translation between DAPR and the research community. Aim 2 is to provide innovative training content to facilitate the use and adoption of DAPR tools. The Resource Core will package and ensure the accessibility of generated data, code, and software, documenting them for user-friendliness. It will also create and share tutorials and quick-start videos on using new software features and accessing harmonized datasets, as well as more general data science, AI and ML tutorials, specific for rehabilitation researchers. The creation and maintenance of online “living guidebooks” on data science and analytics for precision rehabilitation will provide examples in MATLAB, Python, and R. Guidebooks will be regularly updated and initially cover three topics: (1) Basic Data Management and Visualization, (2) Introduction to Data Analysis, and (3) Introduction to AI and Machine Learning for Rehabilitation Scientists. Best practices, datasets, code, software, and tutorials will be embedded, serving as a comprehensive resource for the scientific community. The Resource Core will be a national resource for researchers, offering training in DAPR tools, data science, AI, and ML to advance precision rehabilitation. Leveraging the expertise in training and education of ReproRehab, the DAPR Resource Core will aid with implementing data harmonization, big data analyses, and precision in the rehabilitation research community.