Creating AI-Enabled All-Health Team Data Fabric (CAIDF) Hackathon
May 30 - 31, 2025
Location
Crowne Plaza Chicago Hotel
Address
25 S Halsted St., Chicago, IL 60661
Calendar
Download iCal FileYou are invited! Creating AI-Enabled All-Health Team Data Fabric (CAIDF) Hackathon
Dates: May 30-31, 2025
Location: Crowne Plaza Chicago Hotel, hosted by University of Illinois Chicago, Chicago, IL
Sponsored by: University of Illinois Chicago, University of Iowa and University of Missouri
Supported by: Advanced Research Projects Agency for Health (ARPA-H)
Purpose: This is the first time a large data set is available that includes occupational therapy, physical therapy, speech language pathology, nursing and medical health record data. This hackathon will leverage multidisciplinary data representing over 100,000 patients from three major health systems. Join talented teams of developers, data scientists, clinicians, and healthcare professionals to tackle complex health challenges.
Hackathon focus areas
1. Engineering optimal patient outcomes
Teams working in this area will explore how various clinical interventions and healthcare services impact patient recovery. Ideas that teams might work on include:
Leveraging Diagnosis Related Groups (DRG), and the frequency, sequencing, and intensity of interventions delivered by all of the disciplines to understand impact on outcomes; 2) Length of stay (LOS) evaluation related to number of days and progress on therapy interventions such as stability for Falls patients or ability to feed for Neonatal Intensive Care Unit (NICU) patients; 3) Identify the impact of missed care from one of more of the therapist clinicians, etc.
2. Predictive analytics
Teams will explore how large, complex datasets from multiple care teams can lead to new discoveries and predictive insights for improving patient outcomes. Dive into patient data to identify hidden patterns using advanced Artificial Intelligence (AI) and Machine Learning (ML) techniques.
3. Advanced all-team summarization and visualization
Teams will develop solutions to summarize, integrate, and visualize complex multidisciplinary data. This will both provide clinicians with a clearer picture of patient progress and future care needs and enable patients to understand their hospital journey and manage post-discharge care.
Resources
- Secure data enclave
- Structured health record from data from 2 complex patient populations: (1) Patients who experience a Falls with injury (N= 87,922) and (2) NICU patients transitioning home (N=14,021) from four Midwest academic health systems.
- PCORnet common data common model is available for the structured MD data. Python, R, and other common tools will be available. Additional software tool requests will be considered in advanced of the event.
Register Now!
This event is by invitation only, please make sure to sign up by April 30th if you want to participate. Participation is limited to 100 participants.
Date posted
Apr 9, 2025
Date updated
Apr 9, 2025