CLINICAL DEVICE CONNECTORS

Protocols & Standards

4 of 10

Connect clinical devices and professional peripherals through standards-based local, network and vendor-specific interfaces.

IEEE 11073HL7 / FHIRBLE / GATTASTMUSB / HIDVendor SDKs

Protocols & standards matrix

Standard / protocol	Common usage	Typical Mediloop output	Notes
IEEE 11073	Personal / professional health device communication	Observation, Device, DeviceMetric	Good interoperability baseline for measurements
HL7 v2	Device gateways and legacy enterprise feeds	Observation, DiagnosticReport, ADT-linked workflows	Often seen behind vendor middleware
FHIR R4/R5	Modern device/cloud APIs	Observation, Device, Encounter-linked resources	Best fit for canonical exchange
BLE / GATT	Local wireless devices	Measurements, status, battery, metadata	Usually requires local Agent/mobile bridge
ASTM	POC / analyzer workflows	Results mapped to Observation / DiagnosticReport	Legacy but still relevant in some settings
USB / HID / Serial	Workstation-connected peripherals	Workflow events, identifiers, raw device outputs	Often mediated by Workstation Agent
Vendor SDK / proprietary	Specialty or advanced device functions	Connector-specific canonical model then FHIR	Preserve provenance and versioning

Practical interpretation

Protocol support is about both transport and meaning.

Structural interoperability alone is not enough; map semantics too.

Always preserve device ID, model and firmware metadata when available.

Avoid discarding raw measurements if normalization could change meaning.

MSIS boundary

MSIS is typically used after payload parsing, when you need unit normalization, terminology alignment or provenance-aware reconciliation of clinical values.

Canonical mapping targets

Device identityFHIR Device / DeviceMetric

MeasurementsFHIR Observation

Diagnostic resultsFHIR DiagnosticReport with Observations

WaveformsFHIR Observation or referenced binary/derived assets depending on modality

AlertsEvent / alert stream + workflow events

Operational metadataConnector canonical model with provenance

Illustrative transformation

typescriptCopy

const parsed = parseVendorPayload(payload);
const normalized = await msis.normalize({
  value: parsed.value,
  unit: parsed.unit,
  code: parsed.code,
  sourceSystem: parsed.vendor,
});

return {
  resourceType: 'Observation',
  status: 'final',
  code: normalized.code,
  valueQuantity: {
    value: normalized.value,
    unit: normalized.ucumUnit,
  },
  device: { reference: 'Device/dev-123' },
};