Blog posts

Digital Health Symposium: Advancing Veteran Care through Innovation

Prof. Shehan Hettiaratchy presenting Op RESTORE, NHS England’s veterans physical health and well-being service

On 24th September 2025, the AI for Healthcare Centres and Op RESTORE, NHS England’s veterans physical health and well-being service, hosted a Digital Health Symposium at Imperial College London’s Data Observatory. This brought together clinicians, researchers, veterans’ representatives, and digital health innovators. The event, organised around the NHS Op RESTORE programme, explored how AI and digital tools can transform care for veterans and beyond.

Digital Health for Complex Needs

The day opened with reflections on how digital health is reshaping medicine, surgery, and training. Discussions drew on themes from the Sovereign AI whitepaper, which outlined six key dimensions of sovereignty spanning Data, Models, Compute, Governance, Training, and Legal/Ethics. The framework emphasises a mission-driven, resilient approach to building modular sovereign systems that balance collaboration with independence.

Prof. Aldo Faisal highlighted the potential of initiatives such as Nightingale AI, which aims to transform the NHS’s vast data resources into actionable knowledge access points, as well as the opportunity to build sovereign health foundation models that protect UK data while attracting global investment.

NHS Op RESTORE: Holistic Care for Veterans

The focus then shifted to the NHS flagship programme for veterans: Op RESTORE, which provides specialist care for those with physical health needs following military service.

  • Natalia Sanchez-Thompson (Imperial College NHS Trust) underscored the urgency of improving healthcare, noting that “public satisfaction is at its lowest ever” and that veterans often present with specific needs not well understood in mainstream NHS services. A key innovation has been enabling veterans to self-refer directly into Op RESTORE, bypassing traditional bottlenecks.
  • Neil Brocklehurst (Op RESTORE Data Scientist) shared how AI is being used to support triaging rather than diagnosis. By focusing on characterising specific conditions, the system can free up 20-30% of clinical capacity- a step-change that could be scaled nationally. Importantly, AI triaging allows veterans to self-refer into a streamlined pathway, using insights from over ten years of patient data.
  • Prof. Shehan Hettiaratchy (National Clinical Director, NHS Op RESTORE and Professor of Practice, Imperial College London) spoke about the need for holistic care that is aligned with wider policy and strategy, stressing that the model must be scalable to meet growing demand.

AI4Health Innovation in Action

During the Imperial College London Showcase, Amr Nimer and Dr Bala Kadirvelu presented their work on digital mental health, highlighting Mindcraft, an innovative application designed to support adolescents’ psychological wellbeing. The app demonstrates how AI-driven tools can be used to enhance resilience, provide accessible mental health support, and scale interventions across wider populations. Speakers including George Chewning (US Veterans Administration), Susan Walton (Combined Ops), Ben Robb (Veteran X), Matt Kiernan (Northumbria University), and Hugh Andree (WeServed) provided diverse perspectives on veteran needs, digital tools, and community-driven approaches to care. The event was also attended by Sir Nick Pope, chair of COBSEO, the UK military charities federation and members of the Office for Veterans’ Affairs.

These discussions underscored the importance of both patient feedback and strategic communication in scaling digital solutions responsibly.

Towards Scalable, Resilient Digital Health

A recurring theme throughout the symposium was that no single institution can deliver healthcare innovation in isolation. Success requires integrated, collaborative solutions that bridge policy, research, technology, and frontline care.

The event highlighted the transformative potential of digital health, not only for veterans, but for wider patient populations. Whether through AI-supported triaging, personalised digital twins, or sovereign AI models, the future of healthcare lies in leveraging data and technology responsibly, with patients at the heart.

Learn more about Op RESTORE, The Veterans Physical Health and Wellbeing Service here.

This event was supported by UK Research and Innovation. UKRI Centres for Doctoral Training in AI for Healthcare grant number EP/S023283/1, UKRI AI Centre for Doctoral Training in Digital Healthcare grant number EP/Y030974/1 and The School of Human and Artificial Intelligence. 

 

AI4Health Industry Day 2025

On 24th September 2025, the AI for Healthcare Centres hosted their annual Industry Day at the City and Guilds Building, Imperial College London. The event brought together leading voices from academia, industry, healthcare, and regulation to explore how artificial intelligence is shaping the future of healthcare.

With a packed programme of talks and networking, chaired by Prof. Hamed Haddadi, the day highlighted both the opportunities and challenges in translating AI from research to real-world clinical impact.

AI4Health and Convergence Science

The day opened with a welcome from the AI4Health Director, Prof. Aldo Faisal, outlining the Centres’ vision to train 233 PhD students by 2032 at the interface of AI and healthcare. The introduction also emphasised convergence science – the need to bring together diverse disciplines, sectors, and stakeholders to deliver truly integrated healthcare solutions.

Key initiatives were also showcased, including the launch of Nightingale AI and pioneering work on building a complete digital twin at St Mary’s Hospital emergency department. The message was clear: meaningful change in healthcare requires not just technical innovation, but also partnerships and shared vision.

Morning Highlights: Innovation, Data, and Trust

  • Sultan Haider (Siemens Healthineers) explored the success factors for innovation in healthcare, stressing that real progress begins with acquiring a mandate for change and aligning efforts around a purpose greater than profit: building a healthier society.
  • Alex Zaretsky (British Standards Institute) addressed trust in AI as a medical device, underlining the importance of certification and standards beyond technical files, and the need to bridge deep tech with regulatory frameworks.
  • Oracle & Lewisham Health and Care Partners shared the impact of their Did Not Attend (DNA) Risk Model, which predicts non-attendance at hospital appointments. Between March and August 2024, over 20,000 patients were contacted, reducing missed appointments and saving costs. Their case study showed how machine learning can deliver people-centred improvements in the NHS.
  • Sinead O’Connor (Trinity College Dublin) provided insights into the European Health Data Space (EHDS) and its implications for AI in healthcare. She noted both the opportunities and barriers, highlighting that even patients often struggle to access their own data under current regulations. The EHDS promises a more unified framework for secondary use of health data, with potential to unlock pan-European research and innovation.

Afternoon Sessions: From Interoperability to Foundation Models

  • Jeff Fried (InterSystems) described how the company manages over 1 billion patient records globally, including 60% of NHS data in the UK. He showcased how interoperability and trusted data systems are essential for scaling AI, highlighting partnerships with companies tackling neonatal sepsis and dementia through AI-driven solutions.
  • Paul Comerford (Information Commissioner’s Office) spoke about data minimisation and privacy-enhancing technologies, underscoring the regulatory guardrails that must guide AI adoption.
  • Dimitris Spathis (Google) discussed the potential of foundation models for personal health, and how signals from personal devices can fuel multimodal, fair, and robust AI systems.

AI in Action: From Concept to Impact

The day closed with a showcase of health-tech innovators translating AI research into practice. These examples demonstrated how cutting-edge research is already finding its way into clinics, improving efficiency and patient outcomes.

  • Maria Leiloglou (Enacuity) presented on AI-driven hyperspectral imaging for minimally invasive surgery.
  • Natalie Cooke (SurgeryAI) introduced her company’s mission to empower surgeons with AI tools for precision and workflow optimisation.
  • Boroumand Zeidaabadi (Cardiovolt.ai) presented his company dedicated to turning everyday heart traces (ECGs) into an AI-powered early-warning risk tool that detects silent disease and predicts future health events years before symptoms emerge.

The AI4Health Industry Day 2025 reaffirmed the importance of collaboration across academia, industry, healthcare providers, and regulators. The event not only offered students and partners the opportunity to network but also showcased how the Centres act as a bridge between discovery and deployment.

 

This event was supported by UK Research and Innovation. UKRI Centres for Doctoral Training in AI for Healthcare grant number EP/S023283/1, UKRI AI Centre for Doctoral Training in Digital Healthcare grant number EP/Y030974/1 and The School of Human and Artificial Intelligence. 

AI4Health Spotlight: PhD Researchers Present at MICCAI 2025

We’re thrilled to celebrate the achievements of our AI4Health PhD researchers, who have had their work accepted at the International Conference on Medical Image Computing and Computer-Assisted Intervention (MICCAI 2025), held this year in Daejeon, South Korea. MICCAI is one of the most prestigious global conferences in the field, and our students’ success reflects both their dedication and the strength of the AI4Health programme.

Our CDT researchers represent the next-generation of innovators in AI, driving knowledge and application of AI to healthcare, and it is crucial that their work is disseminated in the leading international conferences and journals in subject-specific or general AI to help understand the state-of-the-art and network with their global peers. The Centres provides support with expert mentoring, with novel connections with patient-oriented charities, NHS Trusts and leading healthcare companies.

Spotlight on our researchers

• Sarah Cechnicka will present her first-author paper at the MICCAI GRAIL workshop. Her work highlights innovative approaches to medical image analysis and was selected for its originality and impact. Read the paper here.

• Connor Daly had his paper, “Towards Markerless Intraoperative Tracking of Deformable Spine Tissue”, accepted to the main conference – an incredible achievement, given the record-breaking 3,667 submissions this year and an acceptance rate of just 9%. His work represents the first real-world clinical application of markerless tracking for orthopaedic and spine surgery. Connor will present in the Oral & Spotlight Session 5 on September 24, 2025.

• Weitong Zhang will contribute both a paper and a presentation on “Multi-Agent Reasoning for Cardiovascular Imaging Phenotype Analysis”. This work advances the use of AI to interpret complex cardiovascular imaging data.

• Vishal Jain will present his research on KidneyGrader: Fine-Grained Tubulitis Scoring Using Weakly Supervised Transformers, as part of both a workshop and conference sessions. His contributions push the boundaries of how AI can support more precise diagnosis in nephrology. This research was conducted in collaboration with Abrar Rashid, a MEng final year student.

• Arthur Lefebvre will also be presenting at the Dehazing Echocardiography Challenge on September 27, having made it into the top 10 teams.

Dr Ahmed Fetit, Senior Teaching Fellow, AI4Health said: “We’re incredibly proud of our PhD students whose research has been accepted at MICCAI and its workshops this year! As one of the world’s leading conferences in the field, MICCAI offers them the chance to share their cutting-edge work with the global medical image computing community. This success is a testament to their creativity and hard work, with the guidance of their supervisors and support from the AI4Health programme.”

We look forward to seeing how their research continues to grow and inspire new collaborations across healthcare and technology.

Learn more about MICCAI 2025: MICCAI Conference website.

This research was supported by UK Research and Innovation. UKRI Centres for Doctoral Training in AI for Healthcare grant number EP/S023283/1, UKRI AI Centre for Doctoral Training in Digital Healthcare grant number EP/Y030974/1 and The School of Human and Artificial Intelligence.

 

      

CHOC and Imperial combine forces to develop paediatric medical AI to transform global healthcare

L-R: Dr Robert Kelly, Prof. Mary Ryan, Prof. Terry Sanger, Prof. Aldo Faisal

We were delighted to welcome our partners Prof. Terry Sanger, Vice President for Research and Chief Scientific Officer and Dr Robert Kelly, Medical Director, CVICU and Director of Research and Academics of the Children’s Hospital of Orange County (CHOC Children’s), part of Rady Children’s Health at our Imperial College London campus this week. Over the past few years, the partnership between Imperial’s Prof.  Aldo Faisal‘s and Prof. Padmanabhan Ramnarayan and the CHOC team has been developing a paediatric version of the AI Clinician technology to provide faster and more accurate treatment recommendations for critically ill children.

Terry and Robert were also welcomed on campus by Imperial’s Vice-Provost (Research and Enterprise) Professor Mary Ryan. The culmination of the two day campus and hospital visit was the signing of an agreement for the two institutions to boost their existing partnership focussed on developing artificial intelligence (AI) for paediatric healthcare.  The partnership has already combined 36,000 anonymised records from CHOC with 20,000 anonymised NHS patient records to create the largest international paediatric dataset of its kind.

L-R: Prof. Terry Sanger (CHOC) and Prof Aldo Faisal (Imperial College London) signing the MoU

Prof. Terry Sanger said: “It can be very hard to understand how doctors think as we doctors have a very different thought process to engineers, or mathematicians, or scientists because we are decision-based – what is the next best step for our patient with the knowledge and tools we have? What I love about working with Prof Aldo Faisal and his AI4Health Centres here at Imperial is that they have a multi-disciplinary team that embeds clinicians alongside data scientists and computer engineers. Together they have gone to a tremendous effort to understand in great depth and with great sensitivity the way in which doctors deliver care to their patients, so they really understand the environment in which AI tools such as the AI Clinician and Nightingale AI will be applied and can tailor the tools accordingly. It’s a very exciting initiative to collaborate on and has great potential to support doctors like me to deliver the best care possible.”

Dr Robert Kelly said: “On this two day visit I’ve been struck by how experienced the entire Imperial team is and how much time and effort has been put into the granular nature of this work which makes these tools incredibly robust and useful for clinicians. What we’ve been shown today takes many, many hours of concentrated and dedicated work. They’re very focussed on making these innovations useful for clinicians and that’s where I see my contribution on behalf of CHOC, I can give insight into how these tools could best support me at the bedside when I want to make the best decisions for my patients and their families.”

Prof. Aldo Faisal, Director of the AI for Healthcare Centres (AI4Health) and Co-Director of Imperial’s School of Human and Artificial Intelligence, said:  “AI is a rapidly developing tool that has proven to be valuable across all areas of scientific research but there is a a particular

AI4Health PhD researchers presenting findings

and urgent need to design and develop AI models that are tailored specifically for the medical field. We are pioneering a global not-for-profit effort to develop the world’s largest healthcare foundation model Nightingale AI, which is designed to leverage vast repositories of electronic patient records, biomedical data, and published medical literature to develop an advanced health-focused AI model. CHOC joining this effort is an absolute game changer and puts the often overlooked care of children at the forefront for the best medical AI.

“While the UK’s health system and data integration puts us in a unique position to lead globally in AI in healthcare science and innovation. The gold-standard for developing and evaluating patient-ready AI technology are multi-centre international trials. This strong UK and US partnership enables us to improve children health globally by building AIs that learn from a much broader range of conditions and medical backgrounds.”