In 2024, we began working on a pre implementation phase for an artificial intelligence (AI) programme that could transform how we detect and prevent cancer and diabetic eye disease (retinopathy). Dr Alan Smith, Consultant in Public Health Medicine, tells us how the programme is progressing and how it will deliver measurable improvements in outcomes for our screening participants.
By Dr Alan Smith, Consultant in Public Health Medicine, National Screening Service
We’ve established an AI and Robotic Process Automation (RPA) Strategic Advisory Committee, to advise us on the strategic direction of how AI and RPA can enhance the effectiveness of our screening programmes - BreastCheck, CervicalCheck, BowelScreen, and Diabetic RetinaScreen. More than just about technology, this is about improving how we detect and prevent disease.
Our goal is to improve outcomes for our screening participants, maintain the highest safety and compliance standards, and enhance administrative efficiencies while ensuring patient rights and privacy are protected.
Building public trust through transparency
We’re developing a comprehensive 4-year AI implementation strategy (2026-2029). We recognise that public confidence will be essential to implement AI.
Our implementation plan will include:
- clear, accessible information about AI's role in screening
- dedicated communication resources explaining AI implementation
- public forums to address questions and concerns
- transparent reporting on AI performance and outcomes
- regular progress updates on implementation milestones.
BreastCheck: Leading the way
BreastCheck will be the first programme to explore the use of AI, serving as the foundation for systematic expansion across all our screening services. AI systems will work alongside radiologists to analyse mammograms and identify subtle patterns that might indicate early-stage breast cancer.
While research into AI applications for other screening programmes continues to emerge, breast screening represents the most mature evidence base for immediate implementation. With over 150,000 mammograms every year in BreastCheck, international evidence suggests AI implementation could help detect more early-stage breast cancers.
The evidence for change in breast screening
Recent international research provides compelling evidence of AI's potential in mammography:
- Swedish MASAI Trial: AI-supported breast screening detected 20% more breast cancers while potentially halving radiologist workload.
- German PRAIM Study: Involving over 460,000 women, this mammography study achieved a 17.6% improvement in breast cancer detection rates.
- Danish Health Authority: Reported 33% reduction in radiologist workload in breast screening while maintaining diagnostic accuracy.
Proven results from the implementation of AI in breast screening internationally includes:
- 20-25% reduction in unnecessary follow-up procedures – reducing potential harms for women and reducing system costs
- potential 15-20% improvement in screening sensitivity and specificity – meaning screening will be better at finding cancer when it is present, and better at reassuring women when nothing is wrong
- 25-30% reduction in time spent reading and reviewing screening images - making the screening process more efficient
- 10-15% increase in early-stage detection of breast cancer – improving treatment options and outcomes.
Diabetic RetinaScreen: Advanced eye care
Diabetic RetinaScreen will follow closely behind BreastCheck as the next programme where AI technology is likely to bring benefit. AI-assisted image analysis for retinal imaging will help in the detection of diabetic retinopathy, potentially preventing vision loss in people with diabetes. The system will analyse retinal photographs with precision that complements clinical expertise.
How reliable will AI be?
The implementation will follow a rigorous, evidence-based approach with comprehensive evaluation procedures. Performance monitoring will be built into the system from the beginning, with regular benchmarking against international best practices.
AI may help to reduce the current limitations of screening by improving detection rates.
Will AI replace human doctors?
No. Our implementation strategy will emphasise ‘human-in-the-loop’ approaches where AI will strengthen and not replace clinical expertise. All AI systems will maintain robust human oversight, with clinicians making final clinical decisions.
What about privacy and data protection?
We take AI ethics seriously. Our implementation plan will prioritise regulatory compliance with the EU AI Act and GDPR requirements, alongside robust ethical frameworks addressing algorithmic fairness, transparency, and appropriate human oversight.
The Department of Health has commissioned HIQA to develop a National Framework for the Responsible and Safe Use of Artificial Intelligence in Health and Social Care. This represents a proactive approach to establishing ethical guidelines before widespread AI adoption.
What this means for our screening participants
The experience for our screening participants will remain the same – they will continue to receive invitation letters, attend the same locations, and interact with the same professional healthcare teams. The difference will be behind the scenes, where AI will work with our teams to make sure that any signs of disease are identified as accurately and quickly as possible.
We’ll provide updates about AI's role in our screening programmes, with clear explanations of how the technology supports clinical decision-making and does not replace it.
Building the foundation for systematic expansion
The successful implementation of AI depends entirely on our Data Ecosystem Roadmap (2025-2030), which will enhance how we govern, manage and use our data to improve our screening programmes and the care of our screening participants. This critical infrastructure will make AI implementation possible and will enable future evidence-based expansion across all current and any future screening programmes.
Partnership excellence and leadership
We’re also part of the HSE AI Implementation Framework Working Group, established in January 2025, developing a comprehensive strategy for AI across Ireland's health service.
Our collaborative effort across the HSE, working in partnership with the Department of Health and the HSE Technology and Transformation team, and aligning with the HSE Digital Health Strategic Implementation Roadmap, will position Ireland among the first EU health systems to integrate AI into national population screening programmes.
This initiative will establish Ireland as a recognised leader in healthcare AI implementation.
Our vision for 2030
By 2030, Ireland will operate a world-class AI-enabled screening system that consistently delivers superior clinical outcomes, serves as a model for international health systems, and can provide the foundation for the next advancement in screening - personalised, risk-stratified screening approaches.
This vision aligns with broader Government objectives around digital transformation, innovation, leadership and healthcare excellence. We’re committed to maintaining the highest standards of clinical quality, ethical practice, and public trust.
The future of cancer screening and diabetic eye screening in Ireland is exciting and filled with the promise of detecting disease earlier and more accurately than ever before. Through careful planning, robust governance, and unwavering commitment to patient safety, we are positioned to lead the world in AI-enhanced healthcare screening.