The pipeline below runs entirely in your browser tab. No network request carries your imaging data at any step.
- 1.DICOM FileYour local file, never uploaded
- 2.CornerstoneOpen-source DICOM rendering library
- 3.WebGLGPU-accelerated pixel pipeline
- 4.Local RenderingDisplayed in your browser tab
- 5.No Server UploadZero data transmission
Processing Pipeline
You select a DICOM file or folder from your device. The browser File API reads it into memory — the data path never leaves the local machine. No background upload is initiated at any point.
The file is handed to Cornerstone, the open-source medical imaging toolkit maintained by the OHIF community. Cornerstone decodes the DICOM container, extracts pixel data, and applies photometric interpretation entirely in JavaScript — no server is involved.
Decoded pixel buffers are transferred to a WebGL canvas. The GPU applies window/level (WW/WC) look-up tables, pseudo-colour maps, and zoom transforms in real time. This keeps the main thread free and avoids any server round-trip for image adjustments.
MPR reconstructions, measurements, annotations, and contours are all computed on your device's CPU and GPU. Results are stored in IndexedDB or browser memory — not on our servers. Your study never leaves the browser context.
Our servers handle authentication, project metadata, and analytics only. Pixel data, DICOM tags, and patient-identifiable information are never transmitted to or stored on our infrastructure at any stage of the workflow.