PSI for Multi-Ligament Knee Reconstruction
Our OJSM-published study validating 3D-printed patient-specific instrumentation for multi-ligament knee reconstruction: the tunnel-placement problem, our CT-driven workflow, clinical accuracy, and how it anchors the Salnus planning platform.
Background
Patient-specific instrumentation (PSI) represents a shift in orthopaedic surgery, from generic, one-size-fits-all jigs toward precision-engineered guides built around an individual patient's anatomy. Instead of asking the surgeon to translate a mental plan into freehand drilling in the operating room, PSI carries that plan physically onto the bone.
Our research team at Salnus developed and validated a workflow for designing and manufacturing 3D-printed surgical guides for multi-ligament knee reconstruction. This study, published in The Orthopaedic Journal of Sports Medicine (OJSM), is the first clinical validation of our PSI technology, and it is the evidence base that underpins the rest of the Salnus planning platform.
Why multi-ligament reconstruction is hard
Single-ligament cases, such as an isolated ACL tear, are demanding but well-rehearsed. Multi-ligament injuries are a different problem. When two or more ligaments are reconstructed in the same knee, the bone tunnels for each graft compete for the same limited volume of bone. A tunnel drilled slightly off-axis for one graft can converge on, or even breach, the tunnel for another. Convergence weakens fixation, risks graft compromise, and in the worst case forces an intraoperative change of plan.
The geometry that keeps these tunnels safely separated is patient-specific. It depends on the size of the femoral condyles, the entry and exit points on the cortex, and the three-dimensional trajectory each tunnel must follow. This is exactly the kind of spatial problem that is difficult to solve freehand and well-suited to pre-operative planning carried over into the case with a physical guide.
Key findings
In the validated cohort, the study demonstrated:
- Improved tunnel-placement precision relative to conventional freehand technique, with guide trajectories matching the pre-operative plan
- Reduced intraoperative decision load, because tunnel entry points and angles were pre-planned rather than estimated under time pressure
- Consistent placement across surgeon experience levels, which is the practical promise of PSI: encoding the plan into the guide narrows the gap between a high-volume specialist and a less experienced operator
- No adverse events attributable to PSI use during the study period
The headline takeaway is not a single accuracy number. It is that a planned trajectory can be reproduced in the operating room with a printed guide, which is the prerequisite for everything downstream.
Clinical workflow
Our PSI workflow runs in four stages, and each stage is now a reusable building block in the wider Salnus pipeline:
- DICOM acquisition. Standard CT imaging of the patient's knee. The cortical detail that CT provides is what makes tunnel-entry planning reliable, which is also why our broader tooling leans on CT for surgical planning. We cover the tradeoffs in AI bone segmentation from CT for surgical planning.
- 3D reconstruction. Automated bone segmentation and surface-mesh generation. This is the same segmentation engine described in AI vs manual CT bone segmentation, repurposed here to produce the bone model the guide is fitted to.
- Guide design. Patient-specific cutting and drilling guides shaped to seat uniquely on the patient's anatomy, with tunnel trajectories chosen to avoid convergence.
- Manufacturing. Medical-grade 3D printing in a biocompatible material, sterilizable for theatre use. The general principles are discussed in 3D-printed surgical guides in orthopaedics.
How it connects to the rest of the platform
This publication is the foundation for our broader vision: pairing AI-powered image analysis with patient-specific surgical planning. The same 3D reconstruction pipeline that prints PSI guides now feeds our automated measurement and grading modules. Alignment work such as mechanical-axis measurement (LDFA, MPTA, HKA) and planning tools like AI osteotomy planning draw on the same anatomical reconstruction step that this study validated.
In other words, the OJSM paper is not an isolated result. It is the proof that the reconstruction-to-plan loop works clinically, which is what lets us extend the same loop into measurement, grading, and planning.
"The transition from academic research to clinical application is not just about technology, it is about building systems that surgeons can trust with their patients."
What this does and does not claim
To be precise about scope: this study validates PSI for multi-ligament knee reconstruction. It does not claim a universal accuracy figure for every joint or every procedure, and it does not replace surgeon judgment. The guide encodes a plan; the surgeon still owns the operation. What the work establishes is that a carefully planned tunnel geometry can be reproduced in theatre, reliably enough to matter in the hardest reconstruction cases.
Read the full paper
The complete study is available as open access on the OJSM website:
This work was conducted by Salnus in collaboration with clinical partners. For academic collaboration inquiries, please contact us.
Reviewed by the Salnus biomedical engineering team.