Using AI to build smarter, more resilient healthcare systems

Through the University of Hertfordshire Integrated Care System Partnership, we are applying artificial intelligence to help health leaders better predict, plan, and respond to the evolving needs of 1.6 million people across Hertfordshire and West Essex.

As the NHS faces mounting pressures – longer waiting times, overstretched services, and a growing burden of chronic conditions – calls for digital transformation have grown more urgent.

A landmark 2024 review by Lord Darzi into the state of the NHS in England concluded that while sectors across the economy had embraced digital innovation, the NHS remained ‘in the foothills of digital transformation’. His findings call for a rapid shift from reactive to preventative care, powered by integrated data and intelligent systems.

From patterns to predictions: modelling future demand

Managing today’s health and care systems is increasingly complex. In Hertfordshire and West Essex, population growth and a rapidly ageing population, with a significant rise expected in the over-85s, are placing additional strains on services.

New data science approaches are at the heart of a plan by the Hertfordshire and West Essex Integrated Care System (HWE ICS) to respond to these challenges by building a new intelligence function that uses data and analytics for effective decision making. The expertise of researchers across the School of Physics, Engineering, and Computer Science at the University of Hertfordshire is key to its delivery.

A team, led by Iosif Mporas, Professor of Signal Processing and Machine Learning, is developing AI tools to forecast future demand and understand the current capacity of the health and care system to meet it. Using state-of-the-art machine and deep learning algorithms, they are creating predictive AI models to analyse patient behaviour, group patients by risk factors, and forecast the likelihood of hospital admission or emergency attendance for the most vulnerable.

Traditionally, health system analytics have looked backwards, describing what has already happened to inform disease management. Our aim is to move towards forward-looking models that use AI to project healthcare demand across different population segments – for example, by age, ethnicity, and risk of admission – and to simulate future scenarios shaped by disease prevalence and lifestyle trends. This will help to increase early detection, prevent ill health and identify at-risk populations.

Professor Iosif Mporas,
Professor of Signal Processing and Machine Learning

A partnership for system transformation

The project is being delivered through the University of Hertfordshire ICS (UHICS) Partnership Programme – a collaboration between the University, the NHS Hertfordshire and West Essex Integrated Care Board, Hertfordshire County Council, Essex County Council, and the Voluntary, Community, Faith and Social Enterprise Alliance.

The UHICS partnership was established to mobilise the university’s academic expertise, research capabilities, and facilities in support of local health system priorities. It is helping to drive long-term improvements in service delivery, digital innovation, and workforce development, ensuring projects like this deliver meaningful impact for the region’s 1.6 million residents.

For each part of the health and care system, the project team will model both ‘do nothing’ scenarios – showing how demand and capacity would evolve if no major changes were made – and more realistic future scenarios shaped by planned improvements.

These ‘what if’ scenarios will provide actionable insights. For example, what would be the impact on hospital admissions if more care were delivered remotely? How could early interventions in communities reduce future demand? And what system responses will be needed to meet these projected demands?

Embedding sustainability through training

Building local capability is key. Through a university-led training programme, HWE ICS analysts will learn advanced data science techniques, working with real ICS data on live transformation projects.

“Our models are designed to be interpretable, so clinicians and managers can trust and use them in real-world planning,” says Professor Mporas. “But we also need skilled analytical teams and leaders to ensure these insights drive real change where it matters – on the front line.”

The project reflects the broader ambitions of the UHICS partnership for system-wide transformation. Jennifer Beard, UHICS Partnership Programme Director, says: “By aligning research with local priorities, we’re not just responding to today’s challenges, we’re shaping a more sustainable future. Projects like this show how our research can improve care, build workforce capacity, and support better outcomes across Hertfordshire and West Essex.”

Find out more about UHICS Partnership Programme.

Professor Mporas

Professor of Signal Processing and Machine Learning

Professor Iosif Mporas is a Professor of Signal Processing and Machine Learning at the University of Hertfordshire. His research interests include:

  • speech and audio processing
  • natural language processing
  • spoken and multimodal human-machine interaction
  • brain signal processing
  • biomedical signal processing
  • healthcare monitoring