SciTransfer
OACTIVE · Project

AI-Powered Software That Predicts Osteoarthritis Progression and Personalizes Treatment

healthTestedTRL 5

Imagine you could feed a computer everything about a patient — their body scans, blood markers, activity levels, even their lifestyle — and it would tell you exactly when their knee or hip joint will start breaking down, and what to do about it. That's what OACTIVE built: predictive computer models that work at every scale, from individual cells to the whole body, to catch osteoarthritis early and slow it down. They even developed augmented reality exercise programs so patients actually stick with their rehab because it feels more like a game than a chore. The goal is to move from "your joint is already damaged, here's a painkiller" to "we see trouble coming, let's prevent it now."

By the numbers
12-30%
Prevalence of osteoarthritis in adults aged 65+
13
Consortium partners across multiple disciplines
7
Countries represented in the consortium
5
Industry and SME partners involved
29
Total project deliverables produced
The business problem

What needed solving

Osteoarthritis affects 12-30% of adults over 65 and increasingly hits younger, active populations — yet current treatment is reactive, waiting until joints are already damaged before intervening. There is no widely available tool that integrates a patient's full biological and lifestyle profile to predict when OA will develop and what personalized action to take. Companies in orthopedic devices, rehab tech, and diagnostics lack predictive intelligence that could shift their products from treatment to prevention.

The solution

What was built

OACTIVE produced working prototypes of personalised predictive models for OA prevention and diagnosis, a computational intelligence infrastructure using machine learning and deep learning, an ontology-based data standardisation framework, validated in vitro OA models linking trauma and inflammatory signaling, and augmented reality rehabilitation tools designed to improve patient engagement and treatment adherence.

Audience

Who needs this

Digital health companies building musculoskeletal monitoring or predictive platformsOrthopedic device manufacturers seeking smart/connected product differentiationAR/VR rehabilitation and physiotherapy technology providersBiotech companies developing OA biomarker diagnosticsHealth insurance companies interested in preventive joint health programs
Business applications

Who can put this to work

Orthopedic medical devices and digital health
SME
Target: Companies developing joint health monitoring devices or digital therapeutics platforms

If you are a digital health company struggling to differentiate your musculoskeletal product — OACTIVE developed working prototypes of personalised predictive models that integrate cell-to-whole-body data with biomarker and behavioral information. These models can predict OA onset and progression for individual patients. Licensing this technology could give your platform a predictive edge that no competitor currently offers, targeting a condition affecting 12-30% of adults over 65.

Physiotherapy and rehabilitation technology
any
Target: Rehab clinics or companies building AR/VR exercise platforms

If you are a rehabilitation technology provider looking to improve patient adherence and outcomes — OACTIVE built augmented reality-powered intervention tools designed in a personalised framework. These AR rehab programs were developed specifically for osteoarthritis patients to make treatment more engaging and boost training adherence. With 13 consortium partners across 7 countries validating the approach, the clinical grounding is strong.

Pharmaceutical and biotech diagnostics
enterprise
Target: Companies developing OA biomarker diagnostics or early screening tools

If you are a diagnostics company seeking better predictive markers for joint disease — OACTIVE created computational intelligence models using machine learning and deep learning that combine biochemical and inflammatory biomarkers with patient-specific data. Their in vitro OA models validated the connection between trauma-induced OA and inflammatory signaling. This could accelerate your biomarker validation pipeline for early OA detection.

Frequently asked

Quick answers

What would it cost to license or integrate this technology?

The project was a publicly funded Research and Innovation Action, so the core IP sits with the 13 consortium partners. Licensing terms would need to be negotiated directly with the coordinator (EDEX, Cyprus) and relevant partners. Costs would depend on the scope of use — a single-market license for the predictive models would differ from a global platform integration deal.

Can this scale to large patient populations or clinical networks?

The models were designed to be patient-specific, integrating data from cell level to whole-body level. The ontology-based data standardisation framework (a dedicated deliverable) was built precisely to handle diverse data sources at scale. However, moving from research prototypes to production-grade clinical deployment would require additional engineering and regulatory clearance.

What is the IP situation — can we license specific components?

As an EU-funded RIA project with 13 partners (5 of which are industry/SMEs), IP is typically shared among contributors per the consortium agreement. The predictive models, AR intervention tools, and computational intelligence infrastructure are separate components that may be licensable individually. Contact the coordinator for the IP ownership map.

Does this meet medical device regulatory requirements?

The project produced working prototypes of personalised predictive models and AR-based interventions, but as a research project it focused on scientific validation rather than regulatory approval. Any company looking to commercialise would need to pursue CE marking (MDR) or FDA clearance depending on the target market. The in vitro validation work provides a foundation for regulatory submissions.

How long before this could be deployed in a clinical or commercial setting?

The project ran from 2017 to 2021 and produced working prototypes with validated computational models. Based on available project data, the technology is at a tested-prototype stage. A commercialisation partner would likely need 2-3 years of additional product development, clinical validation, and regulatory work before market deployment.

Can this integrate with existing hospital IT systems or EHR platforms?

OACTIVE developed an ontology-based framework specifically for data standardisation across diverse sources. This was designed to handle multi-scale patient data integration. While direct EHR integration was not a stated deliverable, the standardisation architecture provides a solid technical foundation for interoperability with clinical systems.

Consortium

Who built it

OACTIVE brought together 13 partners from 7 countries (Belgium, Cyprus, Germany, Greece, Spain, Italy, UK), with a healthy 38% industry ratio — 5 of those partners are SMEs. The mix of 4 universities, 4 research organizations, and 5 industry players means the science was developed alongside commercial perspectives from the start. The coordinator is EDEX Educational Excellence Corporation Limited, a higher education institution based in Cyprus. For a business looking to license or co-develop, the presence of 5 industry-oriented SMEs in the consortium suggests there are partners already thinking about commercialisation pathways and market applications.

How to reach the team

EDEX - Educational Excellence Corporation Limited (Cyprus) — contact via project website or CORDIS portal

Next steps

Talk to the team behind this work.

Want an introduction to the OACTIVE team to discuss licensing their predictive OA models or AR rehabilitation tools? SciTransfer can arrange a targeted meeting with the right consortium partner for your use case.

More in Health & Biomedical
See all Health & Biomedical projects