# Data card (draft)

**Dataset / pipeline name:**  
**Version:**  
**Custodian / owner:**  
**Date:**

## Provenance

- Origin (sites, systems, time range):
- Legal authority for collection / use:
- De-identification status (identified, pseudonymous, de-identified):

## Population and representativeness

- Intended demographic and clinical coverage:
- Canadian / Alberta context notes:
- Known under-representation:

## Schema and quality

- Key fields; units; sampling rate (if signals):
- Missingness and outlier handling:
- Label definitions and inter-rater reliability (if applicable):

## Preprocessing

- Cleaning steps; exclusions; train/val/test split policy:
- Leakage checks performed:

## Limitations

- Environmental conditions not covered:
- Device / sensor constraints:

## Retention and sharing

- Retention period; destruction method:
- Cross-border processing locations:
- Subprocessors:

## Risk and privacy

- Re-identification assessment summary:
- Link to PIA reference ID (if any):