PEPR TREASURE · Work Package 6
AI transplant clinic
AI-based diagnostic and prognostic models for multi-organ monitoring
WP6 optimises patients’ immune monitoring by developing AI and machine-learning tools that directly leverage the ultra-precise assays and biomarkers produced by WP4.
- Graft allocation
- Rejection score
- iBox
- Digital twins
Axes and innovations
- 1
Optimising graft allocation
Scalable AI algorithms propose the best donor/recipient match. Beyond classic compatibility criteria, they analyse immunological, genomic and medical profiles to maximise long-term success, reduce waiting-list mortality and assess the potential utility of a transplant.
- 2
Precision diagnostics and fewer invasive procedures
By combining clinical, biological and histological data, AI computes an individualised rejection-risk score to identify patients for whom a biopsy would be unnecessary. A “digital biopsy” generates quantitative lesion scores (qLesion) for automated, standardised analysis, more reliable than conventional histology.
- 3
Prognosis of post-transplant risks
Powerful predictive models anticipate major long-term complications (graft loss, death, cancers, infections, cardiovascular events). The globally validated iBox system will be enhanced to compute these risk scores.
- 4
Personalisation through “digital twins”
All these predictions are brought together in an interactive software suite for clinicians, consolidating digital twins — a dynamic view of the patient’s health — to minimise exposure to immunosuppressants in low-risk patients and adapt therapies in real time.