Ortho Truss is open-source orthodontic software at the intersection of clinical practice, software development, and research. We're actively looking for partners in all three.
Ortho Truss was built in clinical practice, not a lab. That makes it an ideal training environment for residents — real workflows, real HIPAA requirements, real data architecture. We want to work with program directors and faculty to integrate, validate, and improve it together.
Deploy Ortho Truss in your clinic and give us structured feedback on resident workflows, onboarding, and missing features.
Cephalometric tracings, photo series, or treatment outcome data to train and validate the AI modules under IRB.
Residents and attendings who can formally compare Ortho Truss workflows against existing proprietary systems.
Residents or dental students looking for a software-focused capstone or thesis component — we have open modules.
Tell us about your program and how you'd like to get involved.
Ortho Truss is a full-stack Electron desktop application used in a real clinic. The codebase is complex, well-structured, and actively evolving. Whether you want to ship a specific feature, port it to macOS, or improve the AI pipeline — there is meaningful work here.
The app is cross-platform by design but has only been packaged and tested on Windows. macOS builds are the top request.
Academic and research deployments frequently run Linux. AppImage or Snap packaging would open up university use.
Ceph landmark model training and pathology detection are in progress. Computer vision engineers welcome.
Unit and integration tests are sparse. Building out the test layer is high impact for long-term quality.
Tell us what you want to build or contribute.
Ortho Truss is designed with research in mind from the ground up — every action is logged, all AI models are modular and replaceable, and the local-first architecture makes IRB-compliant data handling straightforward. We're looking for researchers who want to study, validate, or extend it.
Ceph landmark detection and photo classification modules need formal validation against clinical ground truth.
De-identified treatment data for studying treatment duration, compliance patterns, and outcome predictors.
Independent security researchers welcome to audit the HIPAA implementation, encryption, and portal architecture.
The aligner planner module models biomechanics from STL meshes. Orthodontic biomechanics researchers invited.
All data sharing requires a formal data use agreement and IRB approval from the requesting institution. De-identification follows HIPAA Safe Harbor standard. We do not share identified or re-identifiable data under any circumstances.
Tell us about your research and what you need from the platform.
Reach out through the contact forms above or directly via GitHub. Every message gets a response.