SMILI (Simple Medical Imaging Library Interface)
Context
SMILI is not a clinic-ready PACS, and it doesn’t try to be. It’s a collection of imaging utilities and a C++/Qt framework that researchers use when they want to experiment with data formats, segmentation, or visualization. In dental imaging this usually means CBCT or panoramic scans: staff pull the datasets out of a PACS, run them through SMILI to try different segmentation or reconstruction approaches, and then export the results as meshes. The project shows up mostly in academic labs because it’s light, cross-platform, and easy to adapt, even if it lacks the polish of commercial software.
Technical Snapshot (table)
| Area | How SMILI is usually applied |
| Platform | Windows, Linux, macOS; built in C++ with Qt |
| Purpose | Research toolkit for I/O, visualization, segmentation, reconstruction |
| Supported formats | DICOM, NIfTI, Analyze, STL/mesh |
| Interfaces | Command line tools, GUI apps, and library APIs |
| Integration | Often tested alongside Orthanc or dcm4che backends |
| Security | None built-in; depends on OS/network environment |
| License | BSD open-source |
| Scale | Best for research groups, teaching labs, prototypes |
Scenarios
Dental research. A university team segments teeth and jawbone from CBCT scans, exporting STL meshes for CAD pipelines.
Teaching setup. Students are given anonymized image datasets and use SMILI to practice 3D reconstructions.
Prototype workflow. IT staff connect SMILI to an Orthanc server, fetch datasets, test algorithms, and push results back into the archive.
Workflow (admin view)
Build or install binaries; most environments need Qt, ITK, and VTK configured.
Import datasets (DICOM, NIfTI, STL) through the GUI or CLI.
Run segmentation/visualization tasks and save results in mesh or volume formats.
For integration, script data pull/push to PACS nodes like Orthanc.
Apply OS-level access control and backups since SMILI itself doesn’t handle security.
Strengths / Weak Points
Strengths
Small footprint, simple to run on lab machines.
Works with a range of imaging and mesh formats.
Open-source and permissive license, good for experiments.
Adaptable for research and educational use.
Weak Points
Not suitable for production PACS or regulated environments.
Security, auditing, and compliance features are absent.
Community is small; updates can be sporadic.
Advanced use requires programming knowledge.
Why It Matters
Dental faculties and imaging researchers often need something between a PACS and a coding framework. SMILI lands in that middle ground: it can open DICOM data, run basic segmentation, and export meshes without locking users into a closed ecosystem. It doesn’t replace PACS, but it extends what can be done with the data — and in teaching or research, that flexibility is often more valuable than a polished interface.
