Terms used across the JupyterHealth documentation.
- Jupyter
- Project Jupyter. An open source project that builds tools and standards for interactive computing, like Jupyter notebooks, JupyterLab, Jupyter Book, and JupyterHub. JupyterHealth is a JupyterHub subproject, and the Hub and data products like Voilà dashboards are built with Jupyter tools. jupyter.org.
- platform
- The JupyterHealth platform. The collection of tools that JupyterHealth builds to work together, mainly the Exchange and the Hub. See How JupyterHealth works.
- deployment
- A running service that someone hosts with JupyterHealth tools. A deployment might include an Exchange, a Hub, or both. The demo deployment at UC Berkeley has both.
- distribution
- A JupyterHub bundled with configuration, software, and services, so it works as one product for its users. The Hub is a JupyterHub distribution.
- data product
- Something a data scientist creates from health data to share with clinicians or patients, like a dashboard, app, or report. Often built in a notebook and shared with a tool like Voilà or Jupyter Book.
- Exchange
- The JupyterHealth Exchange (JHE). A back-end web service that stores patient-consented health data and serves it through APIs, including a FHIR API. Researchers create studies, patients consent and contribute data, and analysts can query it with the client library. See the Exchange documentation.
- organization
- A group in the Exchange that is used to manage access to data. Researchers, clinicians, patients, and studies belong to organizations. See Access Control.
- Hub
- The JupyterHealth Hub. An opinionated JupyterHub distribution that is set up alongside an Exchange. The Hub uses the Exchange for login and has the client library pre-installed. Researchers and data scientists use it to explore data and build data products for clinicians and patients. Its configuration is public, so anybody can copy and adapt it.
- client library
jupyterhealth-client, a Python package for reading data from the Exchange. Returns observations as pandas DataFrames. See the exchange client documentation.- FHIR
- Fast Healthcare Interoperability Resources. The HL7 standard for exchanging clinical data over a web API. The Exchange serves its data through a FHIR API, so clinical data from an EHR and wearable data can be read the same way. hl7.org/fhir.
- SMART on FHIR
- A standard that lets an outside app log in to an EHR or patient portal and read limited FHIR data. JupyterHealth uses it in two ways: clinicians open dashboards from their EHR, and patients bring in records from their hospital’s patient portal. See Share data products with clinicians and patients and smarthealthit.org.
- Open mHealth
- An open schema standard for mobile and wearable health data such as heart rate, steps, and glucose. JupyterHealth stores wearable data in Open mHealth and IEEE 1752 formats. openmhealth.org.
- IEEE 1752
- The IEEE standard for mobile health data, derived from Open mHealth. standards.ieee.org.
- patient-generated data
- Health data that patients collect themselves, outside a clinic, from wearables, sensors, mobile apps, and surveys. JupyterHealth brings it together with clinical data from an EHR.
- patient-consented data
- Data that a patient has explicitly agreed to share with a specific study. The Exchange only accepts the kinds of data from a patient that they have consented to share. See Access Control for what consent does and doesn’t control.
- study
- A group of patients whose data an organization collects in the Exchange for one purpose. Patients choose what to share with each study they join. See Collect patient data.
- CommonHealth
- An Android app, built by the nonprofit The Commons Project, that patients use to connect health devices and send their readings to the Exchange. CommonHealth is a partner project, separate from JupyterHealth. Google Play.
- Open Wearables
- An open source, self-hosted platform that connects to wearable vendors like Oura and serves their data through one API. The Exchange can pull data from an Open Wearables server. Open Wearables is a separate project from JupyterHealth. GitHub.
- EHR
- Electronic health record. The clinical system a hospital or clinic uses to store patient charts, such as Epic or Cerner.
- MRN
- Medical record number. The ID a hospital or clinic uses for a patient. Dashboards launched from an EHR use it to find the same patient in the Exchange.
- CGM
- Continuous glucose monitor. A wearable sensor that records blood glucose every few minutes. The Exchange docs include a CGM tutorial.
- MCP
- Model Context Protocol. An open protocol that lets AI assistants call tools and read data. The Exchange has an optional MCP server, deployed separately, so AI tools can query health data.