The surgeon's engineering
department
for complex knee reconstruction
Send us the hard case: multi-ligament, revision, complex deformity. We turn the CT into a patient-specific 3D plan. Research use only.
Salnus is a research-use-only, software-only, implant-agnostic orthopaedic AI platform for preoperative surgical planning and clinical decision support.
From CT to Surgical Plan
Example: complex multi-ligament knee reconstruction
Tunnel Optimization & Analysis
AI-optimized tunnel trajectories with bone bridge safety analysis. 3D simulation and orthographic projections for multi-ligament knee reconstruction.
Automated Surgical Report
Complete surgical plan with landmark coordinates, anatomical measurements, protocol comparison, and risk analysis. PDF-ready clinical documentation.
Clinical Intelligence
Platform
Cloud-native platform modules and professional services that standardise your preoperative planning, support surgical decision-making, and enable academic collaboration.
Cloud-Based AI Platform
Secure DICOM viewing, AI-assisted analysis, and clinical reporting in one browser-based platform. No PACS integration or software installation required. Currently in invite-only beta.
DICOM viewer + AI analysis + reporting, one platform.SaaS · DICOM · MPR · Cloud · Invite-OnlyAutomated Surgical Reporting
AI-generated clinical reports from radiological measurements (JSW, MPTA, LDFA, FTA). Standardised PDF output with grading assessment, alignment analysis, and surgical recommendations.
From DICOM to PDF report in seconds.PDF · Reporting · AutomatedOA Intelligence Suite
AI-powered knee osteoarthritis grading. DenseNet-121 architecture trained on OAI dataset. Binary OA detection: 84.1% accuracy. Five-class KL grading: 70.3% accuracy. 21+ experiments completed. 27MB model, browser-inference ready.
Knee OA grading with deep learning.KL-Grade · JSW · OA · Deep Learning · DenseNet-121Geometric Analysis & Measurement
Anatomical angle calculation (MPTA, LDFA, HKA, FTA), landmark detection, and mechanical axis alignment analysis. Interactive landmark placement with AI-assisted detection for streamlined workflow.
Auto angles + landmarks + axis alignment.MPTA · LDFA · HKA · Semi-AutomatedAuto-Segmentation & 3D Modelling
Deep learning bone segmentation from knee CT, trained on a multi-center cohort. Four-bone segmentation across femur, tibia, patella, and fibula. Internal validation; manuscript in submission. Automated CT-to-3D mesh pipeline for surgical visualisation and PSI design.
AI bone segmentation + 3D mesh export.nnU-Net · 4-Bone · CT · STL · 3D MeshMRI Cartilage & Meniscus Segmentation
Deep-learning segmentation of bone, cartilage, and menisci from knee MRI. An nnU-Net model in development on a curated cohort, extending our CT segmentation pipeline to soft-tissue and cartilage assessment. Research stage, manuscript in preparation.
Bone + cartilage + meniscus segmentation from MRI.nnU-Net · MRI · Bone · Cartilage · MeniscusMulti-Ligament Reconstruction Planning
A software-only, automated optimizer for multi-ligament tunnel collision: not manual virtual planning, not hardware. Tunnel-angle optimization, bone-bridge safety analysis, and 3D tunnel visualization, from CT landmarks to a complete surgical plan. Grounded in our OJSM-published research.
Optimized tunnel angles + bone bridge safety analysis.Multi-Ligament · Knee · Tunnel Optimization · 3D PlanningDeformity Analysis & Correction Planning
Varus/valgus angle analysis, osteotomy simulation, and mechanical axis correction planning based on Paley principles. End-to-end preoperative deformity assessment for HTO and DFO candidates.
Mechanical axis restoration with digital precision.Deformity · HTO · DFO · Paley3D Modelling & Surgical Guides (PSI)
3D reconstructions precisely matched to the patient's anatomy. Surgical guide design for HTO, DFO, and specific osteotomies, maximising postoperative accuracy. Clinically validated and published in OJSM.
Patient-specific PSI; maximum axial alignment accuracy.3D Print · HTO · DFO · PSITechnical Consultancy
Elite athlete surgical planning, complex deformity case analysis, and biomechanical modelling for challenging clinical scenarios. Expert-level support from biomedical engineers.
Case-by-case expert biomedical engineering support.Elite Athlete · Complex Cases · BiomechanicsCustom Orthopaedic Software
Bespoke clinical software tailored to your practice needs. From deformity correction algorithms to patient management tools, IP shared, revenue model negotiable.
Your clinical need, our engineering, shared IP.Bespoke · Shared IP · Revenue ShareClinical & Academic Partnerships
Clinical dataset curation, algorithm co-development, and joint publication support for medical imaging research. Collaborative AI model training with shared academic output.
IP sharing, revenue-sharing or licensing model.R&D · Academic · Dataset · IP SharingMedical AI Training & Workshops
Hands-on training for surgeons on digital preoperative planning, AI-assisted analysis tools, and 3D surgical guide design. Congress workshops and institutional training programmes available.
Upskill your team in AI-assisted surgical planning.Training · Workshop · Congress · EducationClinical Evidence
Clinical outcomes achieved by our models using real hospital data.
3D-Printed Patient-Specific Guides for Knee Reconstruction
Our study published in The Orthopaedic Journal of Sports Medicine validated the clinical accuracy of 3D-printed patient-specific instrumentation (PSI) for knee reconstruction procedures, demonstrating significant improvement in intraoperative precision compared to conventional techniques.
Four-Bone Knee CT Segmentation Model
Custom nnU-Net deep learning model trained on a multi-center cohort of annotated knee CT volumes. Four-bone segmentation across femur, tibia, patella, and fibula; internal validation, manuscript in submission. The fibula is rarely segmented in dedicated knee CT work, making this a notable result.
Hip Osteoarthritis AI Analysis
Deep learning-based hip osteoarthritis detection and grading model. Systematic review and meta-analysis of 25 studies completed; original model development in progress.
From Academic Research
to Clinical Application
Every module in our products is grounded in our own research, from peer-reviewed publications to studies currently under review.
Active research portfolio, published, under review, and in development
Academic Partnership
AI for Knee OA grading, systematic meta-analyses, and deep learning model co-development. Clinical dataset curation and technical contribution to joint publication processes.
Custom Orthopaedic Software
Deformity correction planning software and 3D surgical guide design with shared IP ownership for your clinical needs. Revenue-sharing model.
Technical Consultancy
Elite athlete surgical planning, complex defect modeling via 3D printing, and DICOM-based AI infrastructure for your clinical workflow.
Why Salnus
Software, not hardware. Vendor-neutral, not locked-in. In the browser, not the cloud.
Implant-agnostic
Not locked to a single manufacturer. Surgeons keep their implant choice.
Software-only
No hardware, no capital cost, no PACS integration.
Browser-based, client-side
Runs in the browser; patient images stay in the clinic (KVKK/GDPR-aligned).
Peer-reviewed foundation
Grounded in a peer-reviewed OJSM 2026 publication. Research-use-only (RUO).
Request
access to the pilot
Reserve early access to the Salnus orthopaedic AI clinical pilot. Surgeons, radiologists, and institutions can request a spot in the upcoming wave.
Other inquiries: info@salnus.com