CBCTScope
A CBCT viewer built by the radiologist who reads with it.
A complete, local-first CBCT viewer in eight purpose-built reading modes: orthogonal MPR with 3D rendering, a multi-slice grid, curved panoramic with nerve tracing, TMJ, freehand reslice, virtual cephalogram, region growing with airway analysis, and two-volume stitching. The same instrument reads 2D radiographs (panoramic, intraoral, computed and digital radiography), and it is free, open source, and local-first: no account, no cloud, no upload, ever.
No agent is required; CBCTScope is a complete standalone viewer. It is also, as far as I can find, the first with native AI-agent control: an agent connected over MCP opens scans, switches reading modes, sets the window, moves through slices, and takes snapshots, while the scan never leaves your computer.
Board-certified oral and maxillofacial radiologist · 100,000+ studies reported
- Reading modes
- 8 purpose-built
- 2D radiographs
- Panoramic · intraoral · CR · DX
- Standalone
- No agent required
- Agent surface
- 8 MCP verbs, navigation only
- Privacy
- Local-only · loopback-bound · no uploads
- License
- AGPL-3.0, open source
MPR + 3D
Multi-slice grid
Curved panoramic
Three of the eight reading modes, on the synthetic phantom. Click any still for full resolution.
Research use only. Not a medical device. CBCTScope visualizes and navigates volumes. It never produces findings, measurements-as-conclusions, or diagnoses, in the UI or over MCP, and it is not for clinical decision-making.
Agent-native imaging
An agent-native viewer exposes clinical navigation verbs, not code execution. CBCTScope’s agent surface is eight verbs, all navigation or visualization: open a scan, switch mode, set the window, move through slices, snapshot, reset. There is no tool that returns an interpretation. That constraint is the design: the agent can operate the instrument the way an assistant operates the viewer’s controls, and the reading itself stays with the human. Neighboring projects put agents next to imaging in other ways; the repo credits them and maps the differences.
The agent moves the camera. The human reads.
Reading modes
Eight purpose-built modes for oral and maxillofacial reading: orthogonal MPR with 3D rendering, a multi-slice grid, curved panoramic reconstruction with nerve tracing, TMJ with axis-corrected sections for both condyles, freehand reslice, virtual cephalogram, region growing with airway analysis, and two-volume stitching. The reconstruction math is original, implemented from scratch and informed by reading practice, not lifted from any vendor. The same instrument also opens single 2D radiographs (panoramic, intraoral, and computed or digital radiography: DICOM PX, IO, CR, DX) under identical display settings.
Privacy by construction
Scans are read in place from a folder you pick; nothing is copied or uploaded. There is no hosted instance and there never will be one. The server binds to loopback and rejects any request whose Host or Origin is not local, so neither another device on your network nor a hostile page in another browser tab can reach it. The bundled demo mode goes further: it can only ever show a synthetic phantom, so demonstrations and recordings can never surface a real case.
Agent-native instruments
One principle, two domains
Bounded verbs for the agent, commit rights for the human. CBCTScope applies it to reading CBCT volumes; Quoin applies it to text.
See Quoin