AFFECT-EU targets atrial fibrillation detection and stroke prevention through digital, risk-based biomarker screening at population scale.
ROCHE DIAGNOSTICS INTERNATIONAL AG
Global diagnostics leader contributing biomarker expertise and commercial validation to EU cardiovascular and neurological screening research.
Their core work
Roche Diagnostics International AG is the Swiss headquarters of Roche's global in vitro diagnostics division — one of the world's largest manufacturers of diagnostic tests, instruments, and digital health tools. In H2020 research, they contribute industry-grade expertise in biomarker development, diagnostic test validation, and commercial pathways for clinical screening tools. Their participation in EU consortia brings a rare combination: deep laboratory diagnostics capability paired with the scale and regulatory know-how to translate research findings into certified diagnostic products. They specifically focus on population-level screening — identifying disease risk in asymptomatic individuals before clinical symptoms emerge.
What they specialise in
AI-Mind focuses on brain connectivity screening and dementia risk estimation in people with mild cognitive impairment.
AI-Mind applies machine learning, deep learning, and multimodal risk assessment to neurological screening, reflecting Roche's move toward AI-augmented diagnostic tools.
Both AFFECT-EU and AI-Mind involve digital screening architectures designed for broad population deployment rather than individual clinical diagnostics.
Risk stratification appears across both projects — for stroke in AFFECT-EU and dementia in AI-Mind — suggesting this is a core methodological contribution.
How they've shifted over time
In their earlier H2020 project (2020), Roche Diagnostics focused on well-established diagnostic territory: biomarkers for cardiovascular risk, atrial fibrillation detection, and linking screening data to population health outcomes — areas where they have strong commercial product lines. By 2021, their focus shifted toward neurological disease, specifically dementia and mild cognitive impairment, and the methodology expanded significantly to include machine learning, deep learning, and multimodal brain connectivity analysis. This trajectory suggests the organisation is actively positioning its diagnostics expertise inside AI-driven disease detection pipelines, moving from biomarker-only approaches toward integrated computational screening platforms.
Roche Diagnostics is moving toward AI-augmented screening for neurodegenerative diseases, suggesting future collaborations will likely sit at the intersection of diagnostics, machine learning, and early dementia detection.
How they like to work
Roche Diagnostics joins consortia as a participant — never as coordinator — which is consistent with large diagnostics companies that contribute validated industry expertise and commercial development capacity while academic partners lead the research agenda. With 45 unique partners across just 2 projects, they operate inside unusually large consortia, averaging 22+ partners per project, indicating multi-stakeholder public health research rather than tight bilateral collaborations. This suggests they are accustomed to complex multi-country governance structures and contribute a clearly defined, bounded function rather than driving the overall scientific direction.
Their 2 projects have connected them with 45 distinct organisations across 18 countries, a remarkably broad network for a two-project portfolio and a clear signal that they join large European public health consortia with wide geographic representation. No single country dominates the collaboration footprint, consistent with pan-European disease screening research.
What sets them apart
Roche Diagnostics brings something almost no other EU research partner can offer: the direct commercial pathway from research biomarker to certified, market-ready diagnostic test at global scale. Consortia that include Roche gain credibility with regulators, payers, and clinical procurement bodies, because a finding that Roche validates is one that could reach clinics within a realistic timeframe. For projects focused on disease screening at population level, their participation signals genuine translation intent — not just a published result, but a product candidate.
Highlights from their portfolio
- AI-MindA 5-year project (2021-2026) combining machine learning, deep learning, and multimodal brain connectivity analysis for dementia risk — the most technically ambitious of the two and the one most likely to produce IP of commercial value to Roche.
- AFFECT-EUPopulation-scale digital screening for atrial fibrillation tied directly to stroke prevention — a high-impact public health objective aligned with Roche's existing cardiovascular diagnostics product range.