School of Computing and Mathematical Sciences

Spotlight on SLAIDER -- Self-learning AI-based digital twins for accelerating clinical care in respiratory emergency admissions

SLAIDER is a £620k project running from 2023 to 2025 and funded by the EPSRC. To provide more context on the project, the team was asked several questions about SLAIDER and how it could impact real-life clinical care.

I understand that you’ll be using a digital twin to help improve respiratory care – what is the challenge that you are aiming to tackle here, and how will the project tackle it?

From a medical perspective, clinical and non-clinical considerations can affect people in different ways. Information collected in a clinical setting, particularly from patients with respiratory conditions varies a huge amount, particularly over time, and information is not always collected or is missing for various reasons. Artificial intelligence models generally can only perform a narrow range of tasks which can only be applied to help manage patients with respiratory conditions at specific points of their treatment. The digital twin model that we propose in this project will tackle this by creating an approach which can deal with the complex nature of patient care  using ‘a self-adaptive learning strategy’ to simulate and understand the entire process for individual patients.

How will the digital twin help? How will it work, in practice? (I understand that it is a computer model)

The self-adaptive digital twin model will provide valuable insights for clinical decision making by allowing clinical staff to visualise a patient’s specific situation, enabling them to be prioritised and providing information on the general or specific condition of the patient. Additionally the aim is to identify currently unknown influences which may lead to respiratory disease or deterioration of the patients earlier, thus helping to determine any steps that can be taken to improve the situation.

What do you hope to achieve – what is the outcome you are hoping for?

We are hoping to develop an example of a Health Digital Twin (HDT) that can be used in future research as a strong working case to lay the foundation for further development and clinical trials. This, in essence, will allow us to carry out clinical case studies to demonstrate how our AI-based HDTs could improve prioritisation of patients through more rapid diagnosis, disease severity and pre-empting deterioration.

How does this work fit into your broader research activity? And the department’s?

Our self-adaptive HDT should accelerate health research and personalised clinical care by providing a proof-of-concept of this idea. An important contribution from our project to the wider research community are software tools to support future research in this space.

Moreover, our project is a perfect fit for adding values to the University’s strategic vision of research in AI, Data Analytics and Modelling for Resilient, Robust, and Trustworthy data-driven AI systems. This is multi-disciplinary project, bringing together a consortium of experts from AI, medicine and social science, which is imperative for Leicester to be at the forefront of research in AI and Data Analytics and we aim to use it as a starting point for developing further innovative projects.

How innovative is this project – how new is it to use this technology for respiratory care?

This is an approach that has been used successfully in engineering previously, but the idea has not been applied to respiratory care until now. We believe in adapting this technology for use in a clinical setting, it is important to incorporate ethical concerns and ensure that the technology is developed responsibly (e.g. it takes into account potential bias). This will be achieved through various approaches, integration of information from multiple sources (both clinical and non-clinical) and by including corrective ‘models’ within the technology. The technology will then be integrated into a software tool to support clinical staff making decisions so they can offer services and care at personalised level. 

How do you see AI making a difference in the clinical setting?

The recent pandemic showed the importance of effectively using data for making decisions - around national policy but also from a clinical care perspective. Digital and AI-assisted healthcare can help to tackle tough situations, especially respiratory care, where digital twin-approaches could play a key role in a paradigm shift towards advanced data-driven approaches for clinical decision support to help free up resources and reduce costs.

Moreover, digital twins have not yet been explored for modelling respiratory care processes and decision making using routinely collected patient data - to offer personalised care, as we propose in this project.

How pleased you are to receive this funding, and how it will help your work?

As responsible AI experts with societal concerns, we are keen to revolutionise areas where AI can make a difference. Respiratory disease is the third biggest cause of death in England, causing on average 68,000 deaths per year between 2013 and 2019 and over 200,000 emergency hospital admissions in 2021-22, with this number continuing to rise. As the Principle Investigator of this project, I am very much pleased with this opportunity and I strongly believe that our project can significantly contribute to accelerate respiratory healthcare at the NHS within the grounds of aiding accurate and effective clinical decision making, improved patient care and user experience and alleviating pressure at the NHS.

Back to top