AI and Digital Unit
The AI and Digital Unit at LSE Health conducts applied, interdisciplinary research at the intersection of artificial intelligence, digital health, and health systems.
Its work addresses a persistent challenge facing contemporary health systems: the gap between rapid technological innovation and the delivery of real-world, equitable, and sustainable value.
The Unit focuses on the applied side of AI and digital health rather than on technical development alone. While the team brings sufficient technical literacy to engage credibly with developers and data scientists, their distinctive contribution lies in understanding how these technologies interact with governance arrangements, regulatory frameworks, financing and reimbursement mechanisms, organisational cultures, as well as clinical practice. These implementation and real-world application gaps are where many promising technologies falter, and where rigorous policy and social science research can add the greatest value.
Situated within LSE Health, the Unit's work draws on our long-standing strengths in health economics, public policy, governance, regulation, and interdisciplinary research. This enables a whole-of-system perspective on AI and digital health, spanning macro-level policy and regulation, meso-level organisational dynamics, and micro-level clinical practice. Through connecting evidence across these layers, the team is able to identify where problems emerge as well as where their underlying causes lie.
Decision-makers are at the centre of our agenda. Policymakers, regulators, health system leaders, and payers are treated as primary users of our research, in the deliberate sense that this work is designed to support real-world decision-making.
Research Areas
This research area examines how regulatory, legal and policy frameworks shape the development, adoption and oversight of AI and digital health technologies. It also includes work on evidence standards and evaluation frameworks that inform policy and governance decisions.
Recent projects and outputs
- Mapping the regulatory landscape for artificial intelligence in health within the European Union: An empirical study of AI governance challenges and regulatory considerations. Read the journal article.
- Evaluation framework for health professionals’ digital health and AI technologies: Produced with LSE Consulting, this work provides evidence-based policy recommendations on digital health evaluation. Read the report.
- Artificial Intelligence, Intellectual Property, and Human Rights: Mapping the Legal Landscape in European Health Systems: An empirical study of how intellectual property rights and human rights co-exist and collide in the context of AI in European health systems. Read the journal article.
This research area focuses on the real-world translation of digital health and AI tools into clinical and system practice. Our work examines barriers to adoption, integration with clinical workflows, and implications for quality and outcomes.
Recent projects and outputs
- HYPERMARKER :Developing and evaluating AI-supported clinical decision support tools for individualised hypertension treatment while considering implementation and evaluation pathways. Learn more about the project and its outputs.
- Mapping Factors That Affect the Uptake of Digital Therapeutics Within Health Systems: Scoping Review: A literature review mapping what goes into the implementation of digital therapeutics. Read the review.
- A scoping review and expert consensus on digital determinants of health: A review and consensus study on how social, political, and commercial determinants have changed in the digital age, as well as what digital determinants have emerged. Read the review.
- Cities@Heart: Developing and evaluating decision support systems for implementing cardiovascular disease interventions.
The research in this area addresses how the impacts of AI and digital health technologies are assessed, including clinical effectiveness, economic value, and broader system implications.
Recent projects and outputs
- Omission and hallucination prevalence of clinical guidelines in diagnostic large language model outputs: An empirical study investigating how consistent large language models can produce information using identical clinical cases. http://dx.doi.org/10.1136/bmjhci-2025-101959
- Responsible integration of AI in health systems: Using Isambard-AI (the UK’s fastest and most powerful supercomputer, purpose-built for AI research - the 11th most powerful supercomputer in the world, the 6th most powerful in Europe, and the 4th greenest in the world), the team is conducting large-scale evaluations of open-source large language models (LLMs) for clinical diagnostic support, spanning 19 models and over 2,000 clinical cases from three international benchmarks and generating over 5.1 million structured LLM outputs
This area examines the value assessment, financing and economic dimensions of digital health and AI.
Recent projects and outputs
- Building Blocks of Value Creation Within Value-Based Health Systems: A Delphi Study: A consensus study on what system-level indicators are most important to the value creation process. Read the journal article.
- Digital Health Reimbursement Strategies of 8 European Countries and Israel: Scoping Review and Policy Mapping: A policy analysis of what different financing mechanisms exist for digital health technologies. Read the review.
Partnerships and engagements
The AI and Digital Unit at LSE Health works closely with public sector institutions, international organisations, industry and civil society. Partners and collaborators include healthcare providers, policymakers, global organisations, and private sector actors, with ongoing engagement across Europe and internationally. The approach is explicitly collaborative and industry-engaged, while maintaining scientific independence and analytical rigour.
Meet the teamExplore our publications