Orthanc

Orthanc receives a CT scan from the scanner via C-STORE.

A Python script triggers an event when the study is complete.

Orthanc sends the pixel data to an AI inference API.

The AI results (JSON) are converted into a DICOM Structured Report (SR).

Orthanc stores the SR and notifies the radiologist via the PACS.

Install a lightweight Orthanc instance at the rural clinic site.

Configure the 'Orthanc-to-Orthanc' peering via HTTPS/REST.

Local doctors push images to the local Orthanc.

Local Orthanc encrypts and pushes data to the central hospital Orthanc.

Central hospital PACS retrieves the images from the central Orthanc.

Researchers upload DICOM files to a specific Orthanc 'Upload' folder.

Orthanc applies an anonymization profile, replacing PatientID with a random pseudonym.

The software scrubs all private metadata tags (0010, 0020, etc.).

The anonymized dataset is moved to a 'Research' archive.

Researchers download the cleaned data via the Web API.

Consolidate images from multiple different vendors (Siemens, GE, Philips) into Orthanc.

Orthanc normalizes metadata across the different manufacturers.

Use Orthanc as the primary source of truth for all imaging data.

Plug in any DICOM-compliant viewer to access the unified database.

Patient scan is uploaded to Orthanc.

Orthanc DICOMweb plugin serves the image stack.

A web-based 3D viewer (OHIF/Stone) requests the data via WADO-RS.

The surgeon views the interactive 3D model on a tablet over secure Wi-Fi.

Connect Orthanc to an AWS S3 bucket using the S3 storage plugin.

Set a rule: Studies older than 2 years move to S3 Glacier.

Orthanc keeps the index metadata locally for fast searching.

If a study is requested, Orthanc automatically retrieves it from the cloud.

Orthanc receives an incoming study.

A Lua script counts the number of expected slices vs. received slices.

If slices are missing, Orthanc sends an alert to the technician's dashboard.

The study is flagged in the database for manual review.