Interleaved Semantic Evaluation

id: https://chorus4ai.org/

name: CHoRUS

title: Patient-Focused Collaborative Hospital Repository Uniting Standards (CHoRUS) for Equitable AI

description: 'CHoRUS for Equitable AI is a Bridge2AI data generation project developing the most diverse,
  high-resolution, ethically sourced, AI-ready critical care dataset to answer the grand challenge of
  improving recovery from acute illness. The project spans 20 academic centers (14 data acquisition centers)
  and is building a publicly available dataset targeting over 100,000 critically ill patients with multi-modal
  data including structured EHR, waveform telemetry, medical imaging, EEG, and clinical notes. All structured
  data is standardized to the OMOP Common Data Model with additional formats (DICOM, WFDB, OHNLP tokenization)
  and comprehensive metadata schemas. Patient-focused efforts determine ethical and legal approaches to
  manage privacy and bias while accounting for Social Determinants of Health. A visualization and annotation
  environment labels data with targets important for prediction. The project emphasizes skills and workforce
  development for a next generation of diverse academic and community AI scientists through training programs
  and partnerships with AIM-AHEAD. As of August 2025, the dataset covers 14 different hospitals with over
  45,000 unique admissions and includes 50,000 patient admissions from ICU, PICU, and NICU, 1.6 billion
  rows of EHR OMOP data, 7,642 admissions with radiology data, and 23 TB of waveform data.

  '

page: https://chorus4ai.org/

language: en

license: Controlled Access - Data Use Agreement Required (OT2OD032701)

publisher: CHoRUS Consortium / Massachusetts General Hospital

is_tabular: false

keywords:
- CHoRUS
- Bridge2AI
- critical care
- acute illness
- AI-ready dataset
- OMOP Common Data Model
- electronic health records
- EHR
- waveform telemetry
- medical imaging
- DICOM
- EEG
- clinical notes
- OHNLP
- health equity
- social determinants of health
- federated access
- multi-modal data
- high-resolution data
- ethical AI
- trustworthy AI
- workforce development
- data standardization
- privacy preservation
- bias mitigation
- ICU
- intensive care

distributions:
- id: chorus:dist:1
  description: 'Structured electronic health record data in OMOP Common Data Model format (proprietary
    clinical data standard). Contains 1.6 billion rows. Available in secure enclave with controlled access
    at https://chorus4ai.org/. Published OMOP schema metadata available.

    '
  path: https://chorus4ai.org/
- id: chorus:dist:2
  description: 'Waveform telemetry data (23 TB) in WFDB (WaveForm DataBase) format following extended
    PhysioNet schema (binary waveform data format). Available in secure enclave with controlled access
    at https://chorus4ai.org/.

    '
  path: https://chorus4ai.org/
- id: chorus:dist:3
  description: 'Medical imaging data (7,642 admissions with radiology; 1,000 images currently available)
    in DICOM format (Digital Imaging and Communications in Medicine) with comprehensive metadata. De-identification
    in process for larger cohort. Planned controlled access at https://chorus4ai.org/.

    '
  path: https://chorus4ai.org/
- id: chorus:dist:4
  description: 'Clinical notes distributed as OHNLP-tokenized plain text following open source OHNLP schema.
    Stored locally at contributing sites (except tokens). Controlled access planned at https://chorus4ai.org/.

    '
  format: TXT
  media_type: text/plain
  path: https://chorus4ai.org/
- id: chorus:dist:5
  description: 'EEG waveform data in EDF+ (European Data Format) and Persyst formats following open source
    schemas (binary physiological signal formats). Extraction in process, planned for controlled access
    at https://chorus4ai.org/.

    '
  path: https://chorus4ai.org/

purposes:
- id: chorus:purpose:1
  name: Improve recovery from acute illness
  description: 'Answer the grand challenge of improving recovery from acute illness by developing high-resolution
    multi-center datasets as a critical first step towards actionable and trustworthy AI in critical care.
    Address the urgent need for infrastructure to support artificial intelligence and machine learning
    (AI/ML) in critical care settings.

    '
- id: chorus:purpose:2
  name: Create AI-ready critical care dataset
  description: 'Develop a publicly available, AI-ready critical care dataset from more than 100,000 critically
    ill patients while ensuring methods promote privacy, accountability, and clinical benefit. Generate
    the most diverse, high-resolution, ethically sourced dataset for AI/ML applications in acute and critical
    care, expanding AI and Machine Learning to improve recovery from acute illness.

    '
- id: chorus:purpose:3
  name: Establish data standards and tools
  description: 'Unify standards to harmonize multi-modal EHR, waveform, imaging, and text data. Develop
    software and tooling to interact with and extract insight from clinical data in diverse formats. Create
    validated semantic mappings for connecting clinical data in various source formats to international
    standards (OMOP Common Data Model, DICOM, WFDB, OHNLP).

    '
- id: chorus:purpose:4
  name: Promote diversity and health equity
  description: 'Ensure comprehensive sets of patient conditions and clinical treatment strategies with
    appropriate contextual factors such as geographic distance to nearest hospital and Social Determinants
    of Health. Develop the skills and workforce for a next generation of diverse academic and community
    AI scientists through comprehensive training and education programs in partnership with AIM-AHEAD.

    '

tasks:
- id: chorus:task:1
  name: Characterize acute and critical care illness
  description: 'Generate data for ML/AI applications aimed at characterizing acute and critical care illness
    patterns, progression, and outcomes across diverse patient populations and hospital settings.

    '
- id: chorus:task:2
  name: Predict complications in critically ill patients
  description: 'Enable prediction of complications among patients with acute or critical illness using
    multi-modal data including structured EHR, waveforms, imaging, and clinical notes.

    '
- id: chorus:task:3
  name: Measure treatment response
  description: 'Support measurement and analysis of treatment response among critically ill patients through
    high-frequency documentation, medication administration records, and clinical outcomes data.

    '
- id: chorus:task:4
  name: External validation for AI model marketplace adoption
  description: 'Provision a holdout test set accessible for model external validation to aid marketplace
    adoption of AI-developed models for implementation in acute and critical care settings.

    '
- id: chorus:task:5
  name: Label data for prediction targets
  description: 'Utilize visualization and annotation environment to label data with targets important
    for prediction tasks in critical care AI applications.

    '

addressing_gaps:
- id: chorus:gap:1
  name: Lack of diverse multi-center critical care datasets
  description: 'Address the absence of large-scale, diverse, high-resolution multi-center datasets for
    critical care AI/ML by creating a dataset spanning 20 academic centers with 100,000+ critically ill
    patients and ensuring balanced, diverse cohorts through federated access and sampling methods across
    14 data acquisition centers.

    '
- id: chorus:gap:2
  name: Insufficient data standardization in critical care
  description: 'Overcome the lack of unified standards in critical care data by harmonizing multi-modal
    EHR, waveform, imaging, and text data to OMOP Common Data Model and other international standards
    (DICOM, WFDB, OHNLP).

    '
- id: chorus:gap:3
  name: Privacy and bias concerns in clinical AI
  description: 'Address ethical and legal challenges in AI through patient-focused efforts that determine
    approaches to manage privacy and bias while accounting for Social Determinants of Health and performing
    community-facing ethics focus groups to determine what data is appropriate for public sharing.

    '
- id: chorus:gap:4
  name: Limited diversity in AI/ML workforce
  description: 'Bridge the gap in diverse AI/ML workforce through comprehensive educational approaches,
    training programs (including AIM-AHEAD partnership), and cultivation of expertise in lay and scientific
    communities to improve AI literacy and utilization.

    '

creators:
- id: chorus:creator:1
  name: Eric S. Rosenthal
  description: Contact PI/Project Leader, Massachusetts General Hospital (MGH), Director of MGH Neurosciences
    ICU
- id: chorus:creator:2
  name: Azra Bihorac
  description: Principal Investigator, University of Florida (UF)
- id: chorus:creator:3
  name: Ashley Cordes
  description: Principal Investigator, CHoRUS Consortium
- id: chorus:creator:4
  name: Gilles Clermont
  description: Principal Investigator, CHoRUS Consortium
- id: chorus:creator:5
  name: Gari David Clifford
  description: Principal Investigator, CHoRUS Consortium
- id: chorus:creator:6
  name: Barbara J. Evans
  description: Principal Investigator, CHoRUS Consortium
- id: chorus:creator:7
  name: Xiao Hu
  description: Principal Investigator, CHoRUS Consortium
- id: chorus:creator:8
  name: Rishikesan Kamaleswaran
  description: Principal Investigator, CHoRUS Consortium
- id: chorus:creator:9
  name: Yulia A. Levites Strekalova
  description: Principal Investigator, University of Florida (UF)
- id: chorus:creator:10
  name: Parisa Rashidi
  description: Principal Investigator, University of Florida (UF)
- id: chorus:creator:11
  name: Cynthia Rudin
  description: Principal Investigator, CHoRUS Consortium
- id: chorus:creator:12
  name: Ishan Canty Williams
  description: Principal Investigator, CHoRUS Consortium
- id: chorus:creator:13
  name: Andrew Ewing Williams
  description: Principal Investigator, Tufts Medicine
- id: chorus:creator:14
  name: Xiaoqian Jiang
  description: Program Lead, UT Health Science Center (UTHealth Houston)
- id: chorus:creator:15
  name: Morteza Zabihi
  description: Lecturer (Machine Learning Basics), Massachusetts General Hospital
- id: chorus:creator:16
  name: Zhenhong Hu
  description: Instructor (Python and Version Control), University of Florida
- id: chorus:creator:17
  name: Debora Simmons
  description: Lecturer (Ethics of AI in Clinical Practice), UT Health Science Center
- id: chorus:creator:18
  name: Aliyah Geer
  description: Workshop Lead (Data Schemas in Clinical Cloud), Massachusetts General Hospital
- id: chorus:creator:19
  name: Ciera McCrary
  description: Program Manager, Massachusetts General Hospital (MGH)

funders:
- id: chorus:funder:1
  name: NIH Common Fund Bridge2AI Program
  description: 'Funded through National Institutes of Health grant OT2OD032701 (project number 1OT2OD032701-01),
    administered by NIH Office of the Director. Opportunity Number: OTA-21-008. Study Section: Data Coordination,
    Mapping, and Modeling (DCMM). Fiscal Year 2022. Total funding in 2022: $5,880,300 (all direct costs).
    Project dates: September 1, 2022 to November 30, 2026 (with approved no-cost extension). Award notice
    date: September 1, 2022. Assistance Listing Number: 93.310.

    '

instances:
- id: chorus:instance:1
  name: Critically ill patients in ICU settings
  description: 'Individual critically ill patients requiring acute or critical care admitted to intensive
    care units (ICU), pediatric intensive care units (PICU), or neonatal intensive care units (NICU) across
    14 data acquisition centers. As of August 2025, the dataset covers 14 different hospitals with over
    45,000 unique admissions, with 50,000 patient admissions from ICU, PICU, and NICU currently released.
    Target enrollment exceeds 100,000 critically ill patients. Retrospective data collection from patients
    with acute or critical illness.

    '

subpopulations:
- id: chorus:subpop:1
  name: Critically ill patients by hospital
  description: 'Patients distributed across 14 different hospitals within the CHoRUS network spanning
    20 academic centers in the United States, ensuring geographic and institutional diversity in critical
    care settings.

    '
- id: chorus:subpop:2
  name: ICU, PICU, and NICU patients
  description: 'Patients admitted to intensive care units (ICU), pediatric intensive care units (PICU),
    and neonatal intensive care units (NICU) across contributing hospitals, with 50,000 patient admissions
    in the current released dataset.

    '

sensitive_elements:
- id: chorus:sensitive:1
  name: Protected health information - clinical EHR data
  description: 'Complete electronic health records including demographics, diagnoses, procedures, medications,
    nursing documentation, and clinical notes for critically ill patients. Contains protected health information
    subject to HIPAA and institutional privacy requirements. De-identified before release through OMOP
    transformation and privacy scanning tools.

    '
- id: chorus:sensitive:2
  name: Physiological monitoring and EEG waveform data
  description: 'Continuous waveform telemetry from bedside monitors and EEG recordings capturing detailed
    physiological states of critically ill patients. May reveal sensitive health conditions and treatment
    responses. Stored in WFDB format (telemetry) and EDF+/Persyst formats (EEG) with controlled access.

    '
- id: chorus:sensitive:3
  name: Medical imaging data
  description: 'Diagnostic imaging studies in DICOM format from hospital PACS systems. De-identification
    in process to remove embedded patient information while preserving clinical utility. As of August
    2025, 1,000 images available with de-id in process for larger cohort (7,642 admissions with radiology
    data total).

    '
- id: chorus:sensitive:4
  name: Social Determinants of Health data
  description: 'Contextual factors including geographic information (distance to hospital) and social
    determinants of health data. Collected to support health equity research while maintaining patient
    privacy through privacy-preserving transformations.

    '

known_biases:
- id: chorus:bias:1
  name: Academic medical center population bias
  description: 'Data collected exclusively from academic medical centers (14 acquisition centers across
    20 academic institutions), which may not represent community hospital or rural care settings. Patient
    populations at academic centers may differ systematically from the broader critically ill patient
    population in terms of disease severity, demographics, and treatment patterns.

    '
- id: chorus:bias:2
  name: Retrospective data collection bias
  description: 'Retrospective collection from hospital electronic health records introduces potential
    selection biases based on documentation practices, coding patterns, and clinical workflows that vary
    across the 14 acquisition centers. Documentation completeness and accuracy may vary by site, modality,
    and time period.

    '

known_limitations:
- id: chorus:limitation:1
  name: Dataset still actively growing and not yet complete
  description: 'As of August 2025, the dataset covers 14 hospitals with over 45,000 unique admissions
    and 50,000 released (ICU, PICU, NICU); target exceeds 100,000 critically ill patients. EEG extraction
    and full imaging de-identification are still in process. Clinical notes are stored locally at sites
    with only tokens available in the enclave. Cohort coverage varies by data modality and ongoing collection
    continues through November 2026.

    '
- id: chorus:limitation:2
  name: Controlled access restricts open research
  description: 'All data requires signed data use agreement and institutional email registration, limiting
    accessibility for researchers at institutions without established data sharing agreements. Access
    through secure enclave (Azure-based infrastructure) imposes computational and logistical constraints.

    '

confidential_elements:
- id: chorus:confidential:1
  name: De-identified clinical data under controlled access
  description: 'All data distributed under controlled access model requiring institutional email registration
    and signed licensing agreement. Data de-identified before release. Full clinical notes stored locally
    at sites; only OHNLP tokens available in enclave. Data use agreement prohibits re-identification attempts.

    '

acquisition_methods:
- id: chorus:acquisition:1
  name: Structured EHR data via OMOP transformation
  description: 'Demographics, medication administration (dosing time-stamped upon each infusion change
    or dose administration), procedures, nursing flowsheets (high-frequency documentation), and diagnoses
    acquired from electronic health records and standardized to OMOP Common Data Model. Controlled access
    with published OMOP schema metadata. Contains 1.6 billion rows of OMOP data as of August 2025.

    '
- id: chorus:acquisition:2
  name: Clinical notes via OHNLP tokenization
  description: 'Clinical notes extracted from EHR systems and tokenized using OHNLP (Open Health Natural
    Language Processing) toolkit. Stored locally at sites except for tokens. Controlled access planned
    with OHNLP open source schema metadata.

    '
- id: chorus:acquisition:3
  name: Medical imaging via DICOM from PACS
  description: 'Imaging data acquired from hospital PACS systems in DICOM format. De-identification in
    process. Planned controlled access with published DICOM schema metadata.

    '
- id: chorus:acquisition:4
  name: Waveform telemetry via WFDB from bedside monitors
  description: 'Bedside monitor waveform data acquired through gateway and middleware systems. Stored
    in WFDB format with controlled access and published PhysioNet schema (extended) metadata. 23 TB available
    as of August 2025.

    '
- id: chorus:acquisition:5
  name: EEG waveforms from hospital databases
  description: 'EEG recordings from hospital databases in EDF+ and Persyst formats. Extraction in process.
    Planned controlled access with open source EDF+ and Persyst schema metadata.

    '

collection_mechanisms:
- id: chorus:collection:1
  name: Retrospective EHR extraction
  description: 'Retrospective data collection from electronic health record systems at 14 data acquisition
    centers. Data extracted includes demographics, medication administration (dosing time-stamped upon
    each infusion change or dose administration), procedures, nursing flowsheets (high-frequency documentation),
    diagnoses, and clinical notes. Standardized to OMOP Common Data Model.

    '
- id: chorus:collection:2
  name: Waveform telemetry capture
  description: 'Continuous waveform telemetry data captured from bedside monitors through gateway and
    middleware systems at each contributing hospital. Stored in WFDB (WaveForm DataBase) format following
    extended PhysioNet schema. Contains 23 TB of waveform data as of August 2025.

    '
- id: chorus:collection:3
  name: Medical imaging acquisition from PACS
  description: 'Medical imaging data acquired from hospital Picture Archiving and Communication Systems
    (PACS) and stored in DICOM format with comprehensive metadata following DICOM schema. 7,642 admissions
    have radiology data; 1,000 images currently available.

    '
- id: chorus:collection:4
  name: EEG recording extraction
  description: 'Electroencephalography recordings extracted from hospital EEG databases in EDF+ (European
    Data Format) and Persyst formats with metadata following open source schemas. Extraction in process
    as of August 2025.

    '

collection_timeframes:
- id: chorus:timeframe:1
  name: Project data collection period
  description: 'Retrospective and ongoing data collection from September 1, 2022 through November 30,
    2026 (project end date with approved no-cost extension). As of August 2025, data covers 14 different
    hospitals with over 45,000 unique admissions. Target enrollment exceeds 100,000 critically ill patients
    by project completion.

    '

data_collectors:
- id: chorus:collector:1
  name: CHoRUS Data Acquisition Centers
  description: '14 data acquisition centers across 20 academic institutions in the United States responsible
    for extracting and contributing clinical data (EHR, waveforms, imaging, notes, EEG) from their hospital
    systems. Site statuses tracked through GitHub interface and Google Form submissions. Data Acquisition
    sub-team manages extraction per Chorus_SOP standard operating protocols.

    '

sampling_strategies:
- id: chorus:sampling:1
  name: Multi-center federated sampling
  description: 'Federated access enables sampling methods to ensure a balanced and diverse cohort across
    20 academic centers (14 data acquisition centers). Legal framework established for collecting data
    at scale with sampling to ensure comprehensive sets of patient conditions and clinical treatment strategies.
    Community-facing ethics focus groups conducted to determine what data is appropriate for public sharing.

    '

raw_data_sources:
- id: chorus:rawsource:1
  name: Hospital EHR systems at 14 acquisition centers
  description: 'Electronic health record systems at 14 data acquisition centers across the United States
    providing structured clinical data (demographics, medications, procedures, diagnoses, nursing flowsheets)
    in diverse institutional formats prior to OMOP standardization.

    '
  source_description: 'Institutional EHR systems (diverse formats) at 14 academic data acquisition centers
    in the United States, containing structured clinical data for critically ill patients.

    '
- id: chorus:rawsource:2
  name: Bedside monitoring systems (waveform)
  description: 'Proprietary bedside patient monitoring systems at participating hospitals serving as the
    raw source for continuous waveform telemetry data prior to WFDB standardization.

    '
  source_description: 'Proprietary bedside patient monitoring gateway and middleware systems at 14 participating
    hospitals, capturing continuous physiological waveform telemetry for ICU patients.

    '
- id: chorus:rawsource:3
  name: Hospital PACS systems (imaging)
  description: 'Picture Archiving and Communication Systems at participating hospitals serving as the
    raw source for medical imaging data in DICOM format prior to de-identification.

    '
  source_description: 'Hospital Picture Archiving and Communication Systems (PACS) at participating institutions,
    containing radiology imaging data (CT, X-ray, MRI, and other modalities) in DICOM format.

    '
- id: chorus:rawsource:4
  name: Hospital EEG databases
  description: 'Hospital electroencephalography databases serving as raw source for EEG recordings in
    EDF+ and Persyst formats prior to extraction and standardization.

    '
  source_description: 'Hospital electroencephalography recording systems and databases at participating
    sites, containing EEG recordings in EDF+ (European Data Format) and Persyst formats.

    '

missing_data_documentation:
- id: chorus:missing:1
  name: Incomplete modality coverage across sites
  description: 'Not all data modalities are available from all 14 acquisition centers. EEG extraction
    and full imaging de-identification are in process as of August 2025. Clinical notes stored locally
    at sites with only OHNLP tokens available in the central enclave. Cohort coverage varies by data modality
    during ongoing data collection.

    '

raw_sources:
- id: chorus:raw:1
  name: Raw multi-institutional clinical data
  description: 'Unprocessed clinical data from 14 acquisition centers in diverse institutional formats
    including proprietary EHR exports, raw waveform streams from bedside monitors, DICOM images from PACS,
    and clinical notes in free-text format prior to any standardization or de-identification processing.

    '

preprocessing_strategies:
- id: chorus:preproc:1
  name: OMOP Common Data Model transformation
  description: 'All structured electronic health record data standardized to the OMOP (Observational Medical
    Outcomes Partnership) Common Data Model. Ensures interoperability and enables use of OHDSI (Observational
    Health Data Sciences and Informatics) tool stack for analysis. Data from all 14 acquisition centers
    unified through this transformation. Validated semantic mappings for connecting clinical data maintained
    in chorus-mapping repository. Characterization reports returned to contributing sites via CHoRUSReports.

    '
- id: chorus:preproc:2
  name: Clinical note tokenization via OHNLP
  description: 'Clinical notes processed using OHNLP (Open Health Natural Language Processing) toolkit
    for extraction and tokenization. Protects patient privacy while enabling natural language processing
    and analysis of clinical text data. Full notes stored locally at sites; only tokens available in enclave.

    '
- id: chorus:preproc:3
  name: Waveform standardization to WFDB format
  description: 'Waveform telemetry data from diverse bedside monitoring systems standardized to WFDB (WaveForm
    DataBase) format following extended PhysioNet schema. Scripts available in chorus_waveform repository.

    '
- id: chorus:preproc:4
  name: Medical imaging de-identification
  description: 'DICOM imaging data undergoing de-identification process to remove patient identifiable
    information from image metadata while preserving clinical utility and image quality. Privacy scan
    tool (privacy_scan_tool) used for medical records privacy scanning.

    '
- id: chorus:preproc:5
  name: Re-identification limitation transformations
  description: 'Data transformed using approaches that limit re-identification while maintaining analytical
    utility. Multiple preprocessing strategies employed to protect patient privacy across all data modalities
    per HIPAA and institutional requirements. Geocoding via DeGauss (UF-Geocoding tool) for OMOP Location
    entities.

    '

cleaning_strategies:
- id: chorus:cleaning:1
  name: Multi-center data harmonization with validated semantic mappings
  description: 'Data from 14 acquisition centers harmonized through validated semantic mappings and standard
    operating protocols (SOPs). Ensures consistency and interoperability across diverse institutional
    EHR systems and clinical practices. Mappings maintained in chorus-mapping repository with clinical
    validation SOP. Cross-site data quality checks through CHoRUSReports characterization reports.

    '
- id: chorus:cleaning:2
  name: Schema compliance validation across all modalities
  description: 'All data modalities validated against published metadata schemas (OMOP, DICOM, WFDB, OHNLP,
    EDF+, Persyst) to ensure compliance and data quality across the multi-modal dataset. Schema extensions
    documented for OMOP high-frequency nursing flowsheets.

    '

labeling_strategies:
- id: chorus:labeling:1
  name: Visualization and annotation environment for prediction targets
  description: 'Custom visualization and annotation environment developed to label data with targets important
    for prediction tasks in critical care AI applications. Supports labeling for characterizing acute
    illness, predicting complications, and measuring treatment response. Annotation interface for clinical
    expert labeling with quality control through review processes per Chorus_SOP standard operating procedures.

    '

intended_uses:
- id: chorus:use:1
  name: AI/ML model development for critical care
  description: 'Primary intended use is development and training of artificial intelligence and machine
    learning models to characterize acute and critical care illness, predict complications, and measure
    treatment response in critically ill patients across diverse hospital settings.

    '
- id: chorus:use:2
  name: External validation of AI models
  description: 'Provision of holdout test set accessible for model external validation to aid marketplace
    adoption of AI-developed models for implementation in acute and critical care settings.

    '
- id: chorus:use:3
  name: Health equity and disparities research
  description: 'Studies examining health equity, social determinants of health, and disparities in critical
    care outcomes across diverse patient populations and hospital settings. Dataset includes contextual
    factors such as geographic distance to nearest hospital.

    '
- id: chorus:use:4
  name: Educational and training purposes for AI scientists
  description: 'Training and education of next generation of diverse academic and community AI scientists
    through hands-on experience with real-world critical care datasets. Integrated with AIM-AHEAD Bridge2AI
    for Clinical Care Training Program (Cohorts 1 and 2).

    '

discouraged_uses:
- id: chorus:discouraged:1
  name: Clinical decision-making without proper validation and regulatory approval
  description: 'Dataset is for research purposes only. AI/ML models developed should undergo appropriate
    clinical validation, regulatory approval, and institutional review before use in patient care or clinical
    decision-making. Models trained on CHoRUS data must undergo independent clinical validation and regulatory
    approval processes (e.g., FDA clearance) before clinical deployment.

    '
- id: chorus:discouraged:2
  name: Re-identification attempts
  description: 'Attempts to re-identify patients from de-identified data violate ethical principles, data
    use agreements, and legal frameworks established for privacy protection under HIPAA and institutional
    requirements. Re-identification attempts are prohibited by the signed data use agreement.

    '
- id: chorus:discouraged:3
  name: Use without awareness of ongoing data collection limitations
  description: 'As data collection continues through November 2026 and quality assurance processes are
    ongoing, early dataset versions should be used with awareness of completeness limitations and ongoing
    expansion (from 45K to target 100K+ admissions). EEG extraction and full imaging de-identification
    still in process as of 2025.

    '

prohibited_uses:
- id: chorus:prohibited:1
  name: Re-identification of de-identified patients
  description: 'Explicitly prohibited under the data use agreement and applicable law (HIPAA). Any attempt
    to re-identify individual patients from the de-identified data is a violation of the data use agreement
    and may result in termination of access and legal consequences.

    '
- id: chorus:prohibited:2
  name: Redistribution of data outside the data use agreement
  description: 'Redistribution, sharing, or transfer of CHoRUS data to third parties outside the terms
    of the signed data use agreement is explicitly prohibited. Data must remain within the secure enclave
    environment or as explicitly permitted by the agreement.

    '

existing_uses:
- id: chorus:existing:1
  name: AIM-AHEAD Bridge2AI Clinical Care Training Program
  description: 'Subsets of the dataset used in the AIM-AHEAD Bridge2AI for Clinical Care Training Program
    (Cohort 1: 2024-2025, Cohort 2: 2025-2026). Training provides foundational hands-on experience using
    Jupyter Notebooks with Bridge2AI CHoRUS ecosystem, workshops on OHDSI/OMOP common data model and clinical
    AI, and development of practical use cases.

    '
- id: chorus:existing:2
  name: Peer-reviewed research publications
  description: 'Dataset methodology and design documented in peer-reviewed publications including Neurocritical
    Care journal (doi: 10.1007/s12028-024-02007). Dataset used for initial characterization and training
    activities as the full cohort undergoes expansion and quality assurance.

    '

future_use_impacts:
- id: chorus:impact:1
  name: Potential to transform critical care AI research
  description: 'The CHoRUS dataset has potential to become a foundational resource for critical care AI/ML
    research, enabling development of models that could improve patient outcomes in ICU settings across
    diverse populations. External validation capabilities support broader adoption of AI-developed clinical
    tools.

    '
- id: chorus:impact:2
  name: Risk of algorithmic bias propagation
  description: 'AI/ML models trained on CHoRUS data may inherit biases present in clinical documentation,
    treatment patterns, or data collection at academic medical centers. Models developed without appropriate
    bias mitigation could perpetuate or amplify healthcare disparities if deployed in clinical practice
    without careful validation.

    '

distribution_formats:
- id: chorus:format:1
  name: OMOP Common Data Model (structured EHR data)
  description: 'Structured electronic health record data distributed in OMOP Common Data Model format
    enabling use of OHDSI tool stack for analysis. Contains 1.6 billion rows. Available in secure enclave
    with controlled access. Published OMOP schema metadata available.

    '
- id: chorus:format:2
  name: WFDB waveform format (bedside monitor telemetry)
  description: 'Waveform telemetry data (23 TB) distributed in WFDB (WaveForm DataBase) format following
    extended PhysioNet schema. Available in secure enclave with controlled access. Published PhysioNet
    schema (extended) metadata available.

    '
- id: chorus:format:3
  name: DICOM format (medical imaging)
  description: 'Medical imaging data (7,642 admissions with radiology; 1,000 images currently available)
    distributed in DICOM format with comprehensive metadata following DICOM schema. De-identification
    in process for larger cohort. Planned controlled access.

    '
- id: chorus:format:4
  name: OHNLP tokenized format (clinical notes)
  description: 'Clinical notes distributed as OHNLP-tokenized text following open source OHNLP schema.
    Stored locally at contributing sites (except tokens). Controlled access planned.

    '
- id: chorus:format:5
  name: EDF+ and Persyst formats (EEG waveforms)
  description: 'EEG waveform data distributed in EDF+ (European Data Format) and Persyst formats following
    open source schemas. Extraction in process, planned for controlled access.

    '

distribution_dates:
- id: chorus:distdate:1
  name: Initial release date
  description: 'Dataset made available for access beginning September 2022 (project start date). Current
    release (August 2025) includes 50,000 patient admissions with 1.6 billion rows of EHR OMOP data. Ongoing
    expansion through November 30, 2026 project end date.

    '

maintainers:
- id: chorus:maintainer:1
  name: CHoRUS Consortium
  description: 'Multi-institutional consortium managing dataset maintenance including 14 data acquisition
    centers, coordinating teams at Massachusetts General Hospital (lead), University of Florida, UT Health
    Science Center, and Tufts Medicine, plus infrastructure development team. Standards, Data Acquisition,
    and Tooling sub-teams manage ongoing data delivery and quality. Contact: cmccrary@mgh.harvard.edu
    (Ciera McCrary, Program Manager, MGH).

    '
- id: chorus:maintainer:2
  name: CHoRUS GitHub Organization (chorus-ai)
  description: 'Active GitHub organization housing repositories for software, semantic mappings, standard
    operating protocols, and project management. 28 repositories with comprehensive documentation and
    community support. GitHub organization at https://github.com/chorus-ai.

    '

updates:
  id: chorus:updates:1
  name: Ongoing data collection and continuous expansion
  description: 'Dataset updated continuously as data collection progresses at 14 acquisition centers.
    As of August 2025, covers 14 hospitals with over 45,000 unique admissions; current released dataset
    includes 50,000 patient admissions (ICU, PICU, NICU) and 1.6 billion rows of EHR OMOP data. Target
    exceeds 100,000 critically ill patients. Project timeline extends through November 30, 2026 (approved
    no-cost extension). Regular status updates tracked through GitHub project management system.

    '

retention_limit:
  id: chorus:retention:1
  name: Long-term dataset retention per NIH data sharing policies
  description: 'Digital data maintained according to NIH data sharing policies and institutional requirements
    at participating centers. Controlled access model ensures long-term availability for research while
    protecting patient privacy. Funded under NIH award OT2OD032701 with project end date of November 30,
    2026.

    '

version_access:
  id: chorus:versionaccess:1
  name: Controlled access versioned releases
  description: 'Dataset versions accessible through secure enclave with controlled access. Users must
    register with institutional email and sign licensing agreement to access current and future versions.
    Version history maintained as data collection expands from 45K to target 100K+ critically ill patient
    admissions.

    '

extension_mechanism:
  id: chorus:extension:1
  name: Site contribution and GitHub-based development
  description: 'Dataset extended through contributions from 14 data acquisition centers via standardized
    extraction protocols (Chorus_SOP). Software tools and semantic mappings extended through GitHub organization
    (chorus-ai) with 28 active repositories. New sites may join through established data acquisition framework.
    Community contributions to mapping efforts via chorus-mapping repository and clinical validation SOP.

    '

ethical_reviews:
- id: chorus:ethics:1
  name: Institutional Review Board approvals at 14 acquisition centers
  description: 'Retrospective data collection approved through institutional review board processes at
    all 14 data acquisition centers across the United States. Community-facing ethics focus groups conducted
    to determine what data is appropriate for public sharing. Legal framework established for collecting
    data at scale.

    '

human_subject_research:
  id: chorus:hsr:1
  name: CHoRUS Human Subjects Research
  description: 'Retrospective data collection from critically ill patients (ICU, PICU, NICU) approved
    through institutional review processes at 14 data acquisition centers. Community-facing ethics focus
    groups conducted to determine what data is appropriate for public sharing. Legal framework established
    for collecting data at scale. Patient-focused efforts determine ethical and legal approaches to manage
    privacy and bias while accounting for Social Determinants of Health. Project includes expertise from
    law, ethics, health services, biomedical science, engineering, and scientific journal publications
    disciplines. Ethics of AI component addressed through AIM-AHEAD training curriculum (safety, risk,
    and legal considerations; IRB, HIPAA/GDPR compliance for OMOP/FHIR data).

    '
  involves_human_subjects: true

informed_consent:
- id: chorus:consent:1
  name: Retrospective data and waiver of consent framework
  description: 'Retrospective data collection from critically ill patients using waiver of individual
    consent framework approved by IRBs at participating institutions. Community-facing ethics focus groups
    engaged to determine appropriate data for public sharing and guide ethical data governance. Patient-focused
    ethics pillar of CHoRUS project determines ethical and legal approaches to privacy, bias, and Social
    Determinants of Health considerations.

    '

at_risk_populations:
  id: chorus:atrisk:1
  name: Critically ill patients and vulnerable ICU populations
  description: 'Dataset includes critically ill patients (ICU, PICU, NICU) who are vulnerable due to acute
    illness and critical care needs. Pediatric (PICU) and neonatal (NICU) patients represent particularly
    vulnerable populations. Dataset also includes patients with diverse social determinants of health
    and geographic factors. Privacy protections include de-identification, controlled access model, and
    data use agreement restrictions. Community ethics focus groups engaged to protect patient interests.

    '

license_and_use_terms:
  id: chorus:license:1
  name: CHoRUS Controlled Access License with Data Use Agreement
  description: 'Dataset distributed under controlled access requiring institutional email registration
    and signed licensing agreement. Access granted after review and approval process. Participants must
    complete registration form with name, institutional email (not personal), and institution. Once approved,
    users receive email with access instructions to CHoRUS secure enclave. For training program access,
    program administrators assist with licensing. Access request contacts - dbold@emory.edu or jared.houghtaling@tuftsmedicine.org.
    Funded under NIH award OT2OD032701.

    '

ip_restrictions:
  id: chorus:ip:1
  name: Institutional data use agreement and NIH award terms
  description: 'Data use is governed by the CHoRUS licensing agreement and NIH grant terms (OT2OD032701).
    GitHub repositories use MIT License and Apache-2.0 licenses for software components. Clinical data
    ownership remains with contributing institutions subject to applicable law. Re-use outside terms of
    the data use agreement is prohibited.

    '

regulatory_restrictions:
  id: chorus:regulatory:1
  name: HIPAA and human subjects research compliance
  description: 'Dataset subject to HIPAA (Health Insurance Portability and Accountability Act) compliance
    requirements for protected health information. Subject to 45 CFR 46 (Common Rule) for human subjects
    research protections. Institutional data privacy regulations at each of the 14 contributing sites
    apply. NIH Common Fund Bridge2AI program ethical and trustworthy AI requirements must be met. Export
    control regulations may apply for international users.

    '

is_deidentified:
  id: chorus:deid:1
  name: De-identified with controlled access
  description: 'All data de-identified before distribution. EHR data de-identified through OMOP CDM transformation
    and privacy scanning tools (privacy_scan_tool). Clinical notes tokenized via OHNLP toolkit with full
    text remaining local; only tokens available in enclave. Medical imaging undergoing DICOM header de-identification.
    Waveform and EEG data in WFDB/EDF+ formats with controlled access. Despite de-identification, controlled
    access model maintained due to sensitivity of clinical data and re-identification risk.

    '

external_resources:
- id: chorus:resource:1
  name: CHoRUS Project Website
  description: Official project website with dataset overview, team information, project components, and
    access instructions
- id: chorus:resource:2
  name: CHoRUS GitHub Organization
  description: Comprehensive GitHub organization with 28 repositories including software, documentation,
    SOPs, and tooling at https://github.com/chorus-ai
- id: chorus:resource:3
  name: Chorus_SOP Documentation Site
  description: Centralized standard operating protocol documentation with interactive workflow diagrams
    for data extraction and contribution at https://github.com/chorus-ai/Chorus_SOP
- id: chorus:resource:4
  name: NIH RePORTER Project Details
  description: Federal grant information and project details from NIH Research Portfolio Online Reporting
    Tools for grant 1OT2OD032701-01 at https://reporter.nih.gov/project-details/10472824
- id: chorus:resource:5
  name: Bridge2AI Program
  description: Parent NIH Common Fund program supporting AI-ready biomedical datasets across four data
    generation projects at https://bridge2ai.org/chorus
- id: chorus:resource:6
  name: AIM-AHEAD Bridge2AI Training Program
  description: Partnership with AIM-AHEAD for Bridge2AI Clinical Care Training Program (Cohort 1 and Cohort
    2) providing AI/ML training for underrepresented trainees at https://aim-ahead.net/
- id: chorus:resource:7
  name: OHDSI Community and OMOP CDM
  description: Observational Health Data Sciences and Informatics community supporting OMOP Common Data
    Model used for CHoRUS structured EHR data at https://www.ohdsi.org/
- id: chorus:resource:8
  name: Published Research (Neurocritical Care)
  description: Peer-reviewed publication documenting CHoRUS dataset methodology and design in Neurocritical
    Care journal at https://doi.org/10.1007/s12028-024-02007