AI READI Dataset Documentation

Description	ID	Name
Better understand salutogenesis (the pathway from disease to health) in Type 2 Diabetes Mellitus usi...	purpose-001	Understanding T2DM salutogenesis
Establish standards, best practices, and guidelines for collection, preparation, and sharing of medi...	purpose-002	Establishing AI/ML data standards
Address the lack of racial and ethnic diversity in T2DM research by creating a dataset that is tripl...	purpose-003	Addressing demographic inequities in T2DM research

📊

Composition

What do the instances represent?

Instances

ID
instance-001
Name
Individual participants
Description
Individual participants aged 40 and older with and without Type 2 Diabetes Mellitus (T2DM). Target enrollment is 4,000 people, triple-balanced by self-reported race/ethnicity (Asian, Black, Hispanic, White), T2DM severity (no diabetes, pre-diabetes/lifestyle-controlled diabetes, diabetes treated with oral medications or non-insulin injections, insulin-controlled diabetes), and biological sex (male, female). Participants must speak, read, and understand English. Exclusion criteria include pregnancy and type 1 diabetes.
Instance Type
Human participants recruited from three health system sites (University of Alabama at Birmingham, University of California San Diego, University of Washington) between 2020 and 2025.

Subpopulations

Description	ID	Name
Self-reported Asian race/ethnicity, target ~1,000 participants (25% of sample)	subpop-001	Asian participants
Self-reported Black race/ethnicity, target ~1,000 participants (25% of sample)	subpop-002	Black participants
Self-reported Hispanic ethnicity, target ~1,000 participants (25% of sample)	subpop-003	Hispanic participants
Self-reported White race/ethnicity, target ~1,000 participants (25% of sample)	subpop-004	White participants
Participants without diabetes diagnosis, target ~1,000 participants (25% of sample)	subpop-005	No diabetes
Participants with pre-diabetes or lifestyle-controlled diabetes, target ~1,000 participants (25% of ...	subpop-006	Pre-diabetes and lifestyle-controlled diabetes
Participants with diabetes treated with oral medications or non-insulin injections, target ~1,000 pa...	subpop-007	Medication-controlled diabetes
Participants with insulin-controlled diabetes, target ~1,000 participants (25% of sample)	subpop-008	Insulin-controlled diabetes

Distribution Formats

Access Urls	Description	ID	Name
https://fairhub.io/datasets/2	Retinal imaging data distributed in DICOM format (converted from proprietary .fda and .sdt formats f...	format-001	DICOM for imaging
https://fairhub.io/datasets/2	Survey data, clinical lab results, continuous glucose monitoring, environmental sensor data, and oth...	format-002	CSV for tabular and time-series data
https://fairhub.io/datasets/2	Physical activity monitoring data (from Garmin VivoSmart 5) converted from .FIT format to mHealth st...	format-003	mHealth standard for wearable data
https://docs.aireadi.org/	Metadata and data dictionaries provided through REDCap system documentation for survey and study coo...	format-004	REDCap data dictionary

🔍

Collection Process

How was the data acquired?

ID *

https://fairhub.io/datasets/2

Name

AI-READI

Title

Artificial Intelligence Ready and Equitable Atlas for Diabetes Insights (AI-READI)

Description

The AI-READI is a flagship dataset consisting of multimodal data collected from 4,000 individuals with and without Type 2 Diabetes Mellitus (T2DM), harmonized across 3 data collection sites (Birmingham, Alabama; San Diego, California; Seattle, Washington). The dataset was designed with future AI/Machine Learning studies in mind, including recruitment sampling procedures aimed at achieving approximately equal distribution of participants across diabetes severity (triple-balanced by race/ethnicity, biological sex, and T2DM severity), as well as a multi-domain data acquisition protocol (survey data, physical measurements, clinical data, imaging data, wearable device data, environmental sensors, biospecimens) to enable downstream AI/ML analyses that may not be feasible with existing data sources such as claims or electronic health records data. The goal is to better understand salutogenesis (the pathway from disease to health) in T2DM. The study follows FAIR principles and incorporates ethical and equitable data collection and management practices.

Page

https://fairhub.io/datasets/2

Language

Keywords

Type 2 Diabetes Mellitus
T2DM
AI-READI
Machine Learning
Artificial Intelligence
multimodal dataset
harmonized data
multi-site study
salutogenesis
FAIR principles
retinal imaging
continuous glucose monitoring
wearable devices
biorepository
biospecimens
triple-balanced sampling
health equity
Bridge2AI
cross-sectional study

Addressing Gaps

Description	ID	Name
Provide a large-scale, harmonized, multi-site, multi-domain dataset enabling AI/ML analyses not feas...	gap-001	Lack of multimodal T2DM datasets
Address demographic inequities in T2DM research by recruiting equal proportions across four race/eth...	gap-002	Demographic underrepresentation
Create a model for future AI-ready medical datasets through comprehensive metadata, standardized dat...	gap-003	AI-readiness of medical datasets

Creators

Role	Name	ORCID	Affiliation
Contributor	Cynthia Owsley	creator-001	-
Contributor	Aaron Lee	creator-002	-
Contributor	Sally L. Baxter	creator-003	-
Contributor	Christopher G. Chute	creator-004	-
Contributor	Megan E. Collins	creator-005	-
Contributor	Jeffrey C. Edberg	creator-006	-
Contributor	Kadija Ferryman	creator-007	-
Contributor	Michelle Hribar	creator-008	-
Contributor	Samantha Hurst	creator-009	-
Contributor	Hiroshi Ishikawa	creator-010	-
Contributor	Cecilia S. Lee	creator-011	-
Contributor	Alvin Y. Liu	creator-012	-
Contributor	Gerald McGwin	creator-013	-
Contributor	Shannon K. McWeeney	creator-014	-
Contributor	Camille Nebeker	creator-015	-
Contributor	Bhavesh Patel	creator-016	-
Contributor	Sara Jean Singer	creator-017	-
Contributor	Michael P. Snyder	creator-018	-
Contributor	Joseph Manuel Yracheta	creator-019	-
Contributor	Linda M. Zangwill	creator-020	-

Subsets

Description	ID	Name
Includes data not considered sensitive personal health information, available to the public for down...	subset-001	Public Access Dataset
Includes sensitive data accessible by entering into a data use agreement. Contains 5-digit zip code,...	subset-002	Controlled Access Dataset
Biobanked samples stored at UAB Center for Clinical and Translational Science (CCTS), including plas...	subset-003	Biorepository

Sampling Strategies

ID
sampling-001
Name
Triple-balanced recruitment
Description
Recruitment sampling procedures aimed at achieving approximately equal distribution of participants across three dimensions: (1) race/ethnicity (Asian, Black, Hispanic, White), (2) T2DM severity (no diabetes, pre-diabetes/lifestyle-controlled, medication-controlled, insulin-controlled), and (3) biological sex (male, female). This balanced design is critical for developing unbiased machine learning models.
Is Sample
- True
Is Random
- False
Is Representative
- False
Strategies
- Targeted recruitment to balance demographics across race/ethnicity, sex, and diabetes severity
- Wave-based recruitment with monitoring and adjustment through under- and oversampling
- Recruitment from electronic health records screening using ICD-10 codes (R73.09 for pre-diabetes, E11.X for T2DM)
- Personalized invitation letters and emails with REDCap recruitment interface

Collection Mechanisms

Description	ID	Name
Single study encounter per participant at one of three data collection sites (Birmingham, San Diego,...	collection-001	In-person data collection visits
Source population identified by screening electronic health records for patients aged 40+ who had me...	collection-002	Electronic health record screening
Participants recruited in waves to facilitate efficient sampling. Composition and size of each wave ...	collection-003	Wave-based recruitment
Continuous glucose monitoring (Dexcom G6, 5-minute intervals), physical activity monitoring (Garmin ...	collection-004	Home-based wearable monitoring
Blood (53 mL) and urine collected during study visit. Local processing for plasma, serum, buffy coat...	collection-005	Biospecimen collection and biobanking

Acquisition Methods

Description	ID	Name
Self-reported data collected via REDCap interfaces including demographics, medical history, social d...	acquisition-001	Survey and questionnaire data
Height, weight, blood pressure, heart rate, waist circumference, body composition, and other anthrop...	acquisition-002	Physical measurements and vital signs
Multi-device retinal imaging protocol capturing data from Optos California, Spectralis OCT2, Triton ...	acquisition-003	Retinal imaging
Visual acuity and contrast sensitivity under photopic (daylight) and mesopic (dim light) conditions ...	acquisition-004	Visual function testing
Complete blood count (CBC) from fresh whole blood at local CLIA-certified labs. Central lab testing ...	acquisition-005	Clinical laboratory testing
12-lead ECG data collected during study visit using standardized protocols.	acquisition-006	Electrocardiogram (ECG)
Montreal Cognitive Assessment (MoCA) administered to assess cognitive function, relevant to T2DM com...	acquisition-007	Cognitive function testing
Monofilament testing performed to assess peripheral neuropathy, a common T2DM complication.	acquisition-008	Peripheral neuropathy assessment
Dexcom G6 Continuous Glucose Monitor capturing blood glucose measurements (mg/dL) every 5 minutes. D...	acquisition-009	Continuous glucose monitoring
Garmin VivoSmart 5 wearable device capturing number of steps, heart rate, sleep duration (circadian ...	acquisition-010	Physical activity monitoring
Custom-designed environmental sensor (Karalis Johnson Retina Center, UW) capturing ambient temperatu...	acquisition-011	Environmental monitoring
Non-fasting blood (53 mL) and urine collection. Processing includes whole blood for CBC, EDTA plasma...	acquisition-012	Biospecimen collection

Preprocessing Strategies

Description	ID	Name	Preprocessing Details
Data harmonized across three collection sites (Birmingham, San Diego, Seattle) using standardized op...	preproc-001	Data standardization and harmonization	Standardized operating procedures across all three sites, Common protocols and equipment, Centralized data management through REDCap
Retinal imaging data converted from proprietary formats (.fda, .sdt) to DICOM standard for the datas...	preproc-002	Image format conversion	Proprietary retinal imaging formats (.fda, .sdt) converted to DICOM, Wearable device data (.FIT) converted to mHealth standard
Standardized local processing for plasma, serum, and buffy coats using consistent protocols. Central...	preproc-003	Biospecimen processing	Standardized local processing for plasma, serum, buffy coats, Centralized PBMC processing at UAB CCTS, Batch shipping for central lab analyses
Multiple quality control measures including standardized training of study coordinators, equipment c...	preproc-004	Quality control and validation	Standardized training of study coordinators, Equipment calibration protocols, ... (+2 more)

Cleaning Strategies

ID
cleaning-001
Name
Multi-site harmonization
Description
Standardized protocols and procedures across all three data collection sites ensure data consistency and quality. Common equipment, training, and REDCap data management system used to maintain FAIR principles compliance.
Cleaning Details
- Cross-site harmonization procedures
- Standardized equipment and training
- REDCap data management for quality
- FAIR principles implementation

Maintainers

ID
maintainer-001
Name
AI-READI Consortium
Description
Multidisciplinary consortium managing dataset maintenance including data collection sites, coordinating centers, and data governance committees.
Maintainer Details
- University of Washington (lead institution, data coordination)
- University of Alabama at Birmingham (biorepository, data collection)
- University of California San Diego (data collection)
- Data Access Committee (access policies)
- Documentation team (version-specific guides)

Retention Limit

ID

retention-001

Name

Data and biospecimen retention

Description

Digital data maintained according to NIH data sharing policies. Biospecimen retention subject to institutional policies and consent agreements. Finite number of biospecimen samples available for distribution.

Retention Details

NIH data sharing policies govern digital data retention
Biospecimen retention per institutional policies
Consent agreements specify retention terms
Finite biospecimen availability

Sensitive Elements

Description	ID	Name	Sensitive Elements Present	Sensitivity Details
Genomic DNA extracted from buffy coats, blood derivatives, and urine samples stored with potential f...	sensitive-001	Genetic and biospecimen data	True	Genetic sequencing data from buffy coats, Blood derivatives and urine biospecimens
5-digit zip code, detailed race, ethnicity, and sex information available in controlled access datas...	sensitive-002	Geographic and demographic identifiers	True	5-digit zip code, Race and ethnicity details, Biological sex
Past health records, medications, traffic and accident reports available in controlled access datase...	sensitive-003	Medical history and records	True	Past health records, Medications, Traffic and accident reports

External Resources

Description	External Resources	ID	Name
Official project website with overview and resources	https://aireadi.org/	resource-001	AI-READI Project Website
Comprehensive dataset documentation with version-specific guides	https://docs.aireadi.org/	resource-002	AI-READI Dataset Documentation
Dataset repository and download portal	https://fairhub.io/datasets/2	resource-003	FAIRhub Dataset Landing Page
Parent NIH Common Fund program supporting AI-ready biomedical datasets	https://bridge2ai.org/	resource-004	Bridge2AI Program
Federal grant information and project details	https://reporter.nih.gov/project-details/10471118	resource-005	NIH RePORTER Project Details
Policies and procedures for data access	https://aireadi.org/goals/data-sharing	resource-006	Data Sharing Information
Additional dataset documentation and resources	https://doi.org/10.5281/zenodo.10642459	resource-007	Zenodo Archive
BMJ Open publication describing study design and protocol	https://doi.org/10.1136/bmjopen-2024-097449	resource-008	Protocol Publication
Overview of AI-READI approach and significance	https://doi.org/10.1038/s42255-024-01165-x	resource-009	Nature Metabolism Commentary

🚀

Uses

What (other) tasks could the dataset be used for?

Tasks

Description	ID	Name
Enable downstream AI/ML analyses across survey, clinical, imaging, wearable device, environmental, a...	task-001	Enable multi-domain AI/ML analyses for T2DM
Support the development of unbiased machine learning models through balanced data collection across ...	task-002	Develop unbiased AI/ML models
Study disease trajectories and salutogenesis pathways in T2DM through cross-sectional analysis of pa...	task-003	Study T2DM disease trajectories

Intended Uses

Description	ID	Name
Primary intended use is development and training of artificial intelligence and machine learning mod...	use-001	AI/ML model development for T2DM
Research leveraging multiple data domains (imaging, clinical, genomic, wearable, environmental) to u...	use-002	Multi-modal T2DM research
Studies examining racial and ethnic disparities in T2DM outcomes, social determinants of health effe...	use-003	Health equity research
Discovery of novel biomarkers for T2DM progression, complications, and salutogenesis using biospecim...	use-004	Biomarker discovery
Use as an exemplar for future AI-ready medical dataset development, demonstrating best practices in ...	use-005	Model dataset for AI-ready data standards

Discouraged Uses

Description	ID	Name
As enrollment is ongoing, pilot data releases and periodic updates may not have achieved balanced di...	discouraged-001	Uses outside pilot phase scope
Dataset is for research purposes. Any AI/ML models developed should undergo appropriate clinical val...	discouraged-002	Clinical decision-making without validation
Attempts to re-identify participants from de-identified data violate ethical principles and data use...	discouraged-003	Re-identification attempts

License And Use Terms

ID

license-001

Name

Creative Commons Attribution Non-Commercial

Description

Public access data distributed under Creative Commons Attribution Non-Commercial (CC BY-NC 4.0) license. Permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited, appropriate credit is given, any changes made indicated, and the use is non-commercial. Controlled access data requires data use agreement. See http://creativecommons.org/licenses/by-nc/4.0/ for full license terms.

License Terms

Proper citation required
Non-commercial use only for public data
Derivative works permitted with attribution
Changes must be indicated
Controlled access data requires separate data use agreement

Motivation

Composition

Collection Process

Uses

Distribution

Maintenance

Human Subjects