Minerva: Patient-Specific Biomechanical Planning for Knee Arthroplasty
Overview
Minerva is an automated analysis platform that turns a routine pre-operative knee CT/MRI into a personalized biomechanical model of the patient’s joint. Rather than reporting bone geometry alone, it characterizes the living knee — bone quality, cartilage, menisci, ligaments, and muscle — and predicts how that joint will behave through the activities a patient actually performs, both as it is today and after a planned implant.
What it does
- Tissue-level mapping. Every structure is measured directly from the imaging at the point level — not as a single average. Each point on a ligament, cartilage surface, or region of bone is assessed relative to its neighbors, so the model identifies local weak spots, loose fibers, and low-density bone that a whole-structure average would hide.
- Dynamic joint behavior. Minerva reconstructs how the femur, tibia, and patella move and load through a range of real activities (walking, deep knee bend, stair descent, drop landing), reporting medial/lateral joint gaps, ligament tension, varus/valgus alignment, and the load required to keep the joint balanced through flexion.
- Surgical decision support. It quantifies clinically actionable questions directly: How much correction does osteophyte removal actually gain? Where does the joint line sit? Will the ligaments be balanced or over-tensioned after the planned alignment and implant? Fracture-risk and bone-quality findings are aggregated over regions, not single noisy voxels, so flags are robust.
- Implant-state simulation. The planned components are placed into the model so the surgeon sees the predicted post-operative joint — ligament attachments, balance, and stability — before the operation.
Minerva is not yet cleared for clinical use and is for research purposes only.