This page describes the structure of Cloud Healthcare API paths and operations
and how they can be used to access and manage data.

## Overview

Healthcare data in datasets and data stores can be accessed and managed using a
REST API that identifies each data store using:

- A Google Cloud project
- A Google Cloud location
- The dataset ID
- The data store type
- The data store ID

The API also implements modality-specific
standards for access that are consistent with industry standards for that
modality.

## Administrative operations

Administrative operations are available for datasets and all data stores. They
primarily consist of creating, reading, updating, and deleting (CRUD) datasets
and data stores. Administrative operations are consistent with most
Google Cloud (Google Cloud) APIs and do not require any adherence
to specific modality standards.

Examples of administrative operations include:

- Creating, deleting, getting, listing, and patching datasets and data stores
- Setting, getting, and testing IAM permissions

## Resource names

A resource name consists of, at minimum, a project ID and a location. It can
extend to include a dataset, data store, and any of a data store's child
resources.

The format for a resource name for a data store that resides within a
Cloud Healthcare API dataset looks like this:

`/projects/PROJECT_ID/locations/LOCATION/datasets/DATASET_ID/DATA_STORE_TYPE/DATA_STORE_ID`

For example, the resource name for an HL7v2 store called `clinical-store1` looks
like this:

`/projects/my-project/locations/us-central1/datasets/my-dataset/hl7V2Stores/clinical-store1`

This resource name shows a project called `my-project` in the
`us-central1` region. The project contains a dataset called `my-dataset`, and
the dataset contains an HL7v2 store called `clinical-store1`.

Operations on a location, dataset, data store, or any of a data store's child
resources all require that a resource name is provided either in
the REST path or the gRPC request.

## Modality paths for modality-specific operations

Operations that access data in a modality-specific data store use a request path
that consists of two pieces: the resource name (for identifying the data store
to access), and a modality path (for identifying the actual data to retrieve).

### FHIR resource modality paths

For example, the full request path for reading a specific FHIR Patient
resource using the patient's ID might look like the following:

`RESOURCE_NAME/resources/Patient/PATIENT_ID`

with `/Patient/PATIENT_ID` being the modality path
(structured according to the FHIR standard) for the Patient resource whose
identifier is specified by `PATIENT_ID`.

### DICOMweb modality paths

DICOMweb requests to retrieve all studies for a given patient would look
like this:

`RESOURCE_NAME/dicomWeb/studies?PatientName=PATIENT_NAME`

As another example, a request to retrieve all instances in a given study
and series would look like this:

`RESOURCE_NAME/dicomWeb/studies/STUDY_UID/series/SERIES_UID/instances`

A request to retrieve an instance would look like this:

`RESOURCE_NAME/dicomWeb/studies/STUDY_UID/series/SERIES_UID/instances/INSTANCE_UID`

In all of these examples, the modality path specification is consistent with
the DICOMweb standard path structure.