FHIR & HL7 bridges¶

The imaging ↔ EHR seam: turn a DICOM instance into a FHIR ImagingStudy, and read or build HL7 v2 messages — so an app or agent can move a study between the imaging world and a clinical system without a separate mapping layer.

DICOM → FHIR ImagingStudy¶

from pydcm import fhir

study = fhir.imaging_study("CT0001.dcm")   # one instance → a FHIR R4 ImagingStudy
study = fhir.imaging_study("study_dir/")   # …or a whole study — every series and
                                           # instance aggregated into one resource
study["resourceType"]                       # "ImagingStudy"
study["numberOfSeries"], study["numberOfInstances"]
study["series"][0]["instance"]

Point it at a single file or a directory / list of files: a study folder is aggregated into one ImagingStudy with every series and instance counted. The study / series / instance hierarchy is mapped from the DICOM headers, and subject.reference is an external Patient/<PatientID> reference (resolve it through a FHIR Patient endpoint). Serialize the dict with json.dumps to hand it to any FHIR consumer.

HL7 v2 — parse¶

from pydcm import hl7

segments = hl7.parse(message_text)         # list of segment dicts

HL7 v2 — build an ORU^R01 result¶

oru = hl7.build_oru(
    config={...},          # sending/receiving application + facility
    context={...},         # patient + order identifiers
    observations=[...],    # OBX result rows
)                          # → ER7 string ready to send back to the HIS

These bridges are pydcm's own API over the native FHIR / HL7 engines — there is no third-party Python FHIR or HL7 library in the loop.