ACLART couples machine learning with in silico finite-element simulation to forecast the clinical outcomes of anterior cruciate ligament reconstruction — before the surgeon makes the cut.
A three-year study aligning patient clinical data, surgical records, and biomechanical simulation into a single predictive tool.
ACLART intersects anterior cruciate ligament reconstruction with AI machine learning and in silico simulation, proposing a predictive model for ACLR clinical outcomes. Each patient’s clinical and surgery data feed into machine-learning algorithms and computational simulation — delivering insights to improve patient selection, surgical decision-making, and pre- and post-operative management.
The innovation sits in coupling AI with finite-element analysis, so decisions the surgeon makes on the table are informed not just by clinical data — but by the biomechanics that follow them.
Extending from the core AI + FEA integration into routine orthopaedic practice.
Subjective failure risk at patient-specific level during pre-op discussion.
Tailored to graft choice, tunnel orientation, and fixation devices.
Identify risk factors before the surgical team commits to a plan.
Large-scale analysis designed to outlive the funded phase.
Simulations that expose graft / bone / boundary interactions.
Peer-reviewed publications, conferences and a national SPOT seminar.
Following TRIPOD and the Guidelines for Developing and Reporting ML Models in Biomedical Research.
Pre- and post-op records, surgical reports, follow-up assessments and MRI imaging from Portuguese centres and Scandinavian orthopaedic registers.
Outlier removal, normalization, one-hot encoding, feature selection and dimensionality reduction to extract patterns that influence ACLR outcomes.
Logistic regression, SVM, random forests and neural networks cross-validated against k-fold and leave-one-out baselines.
Finite-element analysis of graft, bone and fixation devices validated against in vitro experimental knee replicas.
Independent datasets, external validation, a point-of-care clinical calculator, peer-reviewed publication.
Surgical medical · AI machine learning · in vitro biomodelling · in silico biomodelling.
KOOS QOL as the primary regression target with TALS, LKSS and IKDC-SKF evaluated via an ensemble of regression algorithms. End goal: an in-clinic risk calculator for point-of-care outcome discussion with patients.
Stakeholder coordination, SPOT-organised national seminar, and a research protocol for surgeons to contribute ACLR data.
Clinical data from three Portuguese hospitals and Swedish, Norwegian and Danish registers — 1000+ records in hand.
Experimental knee replica, synthetic composite bones, finite-element models replicating experimental behaviour.
Correlating clinical and biomechanical variables to predict revision probability and stratify patients into risk categories.
Senior researchers in medicine, physics, biomechanics and machine learning — with doctoral and master’s students.
Peer-reviewed journals and national / international conferences.
Logo placeholders — to be replaced with supplied marks.