Radiology Report Extractor

Extract the patient, modality, body part, clinical indication, findings, impression, and measurements from a radiology or imaging report PDF. Extraction only.

The full guide: Modality, findings, and impression in a radiology report

How should we start?

Build with Talonic

Need to scale? Create an API key, then run this from your own code or an agent.

Create an API key

Free account required

Start with a document

Upload a file or pick a sample, and see the fields come back.

No signup · nothing stored

Questions about Radiology Report Extractor.

What does the radiology report extractor read?

The accession or report number, the study date and the report (signed) date, the patient name, identifier (MRN), date of birth, and gender, the ordering physician and the interpreting radiologist, the imaging modality (X-Ray, CT, MRI, Ultrasound, PET), and the body part examined.

Are the findings and impression captured?

Yes. The clinical indication, the technique, the comparison to prior studies, the findings (the objective description), the impression (the radiologist’s conclusion), and the recommendations each come back as their own field, reported as written. The tool does not interpret the images or provide a diagnosis.

Are discrete findings and measurements returned as tables?

Yes. A findings details table returns each finding id, anatomical location, description, and severity, and a measurements table returns each measurement type, value, unit, and anatomical reference. The report status, observation code (LOINC), contrast-used flag, and radiation dose also come back.

How is a radiology report different from an operative report or progress note?

A radiology report captures an imaging study’s findings and impression. An operative report is the surgeon’s account of a procedure, and a progress note records a clinical encounter (SOAP). For those, use the operative report extractor or the progress note extractor.

Is it safe to upload a report that contains patient data?

The report is processed via the Talonic API for extraction only, is not retained for training, and is not shared. Handle identified PHI with care and redact where you can; for BAA-covered work, ask about the Talonic platform. The field set follows HL7 FHIR R4 DiagnosticReport and Observation resources with LOINC codes. PDF only, up to 10MB and 100 pages.

Doing this to one file, or to ten thousand?

The tool reads a single document. The platform reads the whole estate once and keeps it queryable — the same engine, with a memory.

See document data extraction if you run this for clinical and billing teams, or the extraction API if you are building it in.