BreastScreening-AI The First Multimodal GenAI Platform Delivering Precision Breast Cancer Diagnosis
- Coordinator
- Sensiperception Lda, Portugal
- Runs
- 1 July 2026 to 31 December 2027
- Status
- Running
- EU funding
- €490k
- Total cost
- —
- Partners
- 1 organisations in 1 countries
- Type
- Coordination and Support Action
- Grant agreement
- 101310071
What the project is doing
In the project's own words, from its summary on CORDIS.
BreastScreening-AI is a multimodal, explainable GenAI platform designed to improve accuracy, speed, and reliability in breast cancer diagnostics. Built by SensiPerception, a deep-tech spin-off from Instituto Superior Técnico, the system integrates mammography, ultrasound, and MRI into a single predictive pipeline using cross-attention deep-learning models trained on a large, expert-annotated multicentre dataset. Unlike existing single-modality AI tools, it delivers real-time, BIRADS-aligned assessments enriched with adaptive, user-tailored explanations, supporting both junior and senior radiologists.
The solution addresses persistent diagnostic challenges - false positives up to 40%, false negatives around 10%, and growing radiologist shortages - while operating seamlessly within PACS environments without requiring new hardware. Retrospective validation on >15,000 patient cases showed a 50% reduction in false positives, 60% reduction in false negatives, a fourfold increase in radiologist throughput, and a reduction of diagnostic time from 11 minutes to under 3.The EIC Pre-Accelerator will enable BreastScreening-AI to advance from TRL 4 to TRL 6 through clinical-workflow integration, multicentre pilots, and regulatory-aligned validation supporting CE-IVDR Class IIb and FDA pathways.
In parallel, the project will consolidate its IP position, strengthen business readiness, and prepare for scale-up through Siemens Healthineers’ syngo.via Frontier ecosystem. By combining deep clinical reasoning with trustworthy, human-centred AI, BreastScreening-AI aims to become the European reference solution for early, accurate, and explainable breast cancer diagnostics.
Topics
Who is involved
| Organisation | City | Country | Role |
|---|---|---|---|
| Sensiperception Lda | Coimbra | Portugal | Coordinator |
Source: CORDIS, European Commission. Data of 5 Nov 2026, reused under the Commission's reuse policy. View on CORDIS · Grant DOI