We are advancing artificial intelligence (AI) and data science in health and medicine through computational methodologies that encompass data processing, management, visualisation, modelling, systems, and human–computer interaction, thereby strengthening health research, clinical decision-making, and therapeutic practice.
Our research spans across multidisciplinary research
The field of AI-enabled healthcare is entering a transformative era, driven by advances in multimodal AI that enable the development of multimodal foundation models. These models, characterized by their ability to learn from large-scale, heterogeneous data, are highly generalizable across tasks and hold substantial potential to reshape clinical practice. The multimodal data used to train these models span diverse sources, including medical imaging, omics, clinical text, physiological signals, and video. Our research focuses on harnessing the rich information embedded in these data to address critical challenges in healthcare, such as early disease detection, personalized treatment planning, and clinical decision support, ultimately aiming to improve patient outcomes through scalable, data-driven solutions.
This research advances scalable, multimodal AI solutions that enhance early diagnosis, enable personalized care, and support clinical decision-making, ultimately improving patient outcomes and healthcare efficiency.
Professor Jinman Kim, Professor Zhiyong Wang, Professor Joseph Davis
The success of modern biomedical and health applications strongly depends on the effectiveness and robustness of accessing and integrating multiple data sources. Biomedical and health data comes in various forms, ranging from central databases, via CSV files and spreadsheets to unstructured formats such as texts and PDFs. In our research, we are investigating the integration of large language models and agentic data engineering pipelines in order to augment structured data such as from CSV files and databases with unstructured data in an efficient way. Another strength of our research expertise is the integration of biomedical data in various storage solutions, ranging from relational databases, to NoSQL document stores, key-value stores and graph data systems. We investigate here too the role of modern AI-supported systems for automated schema integration and entity matching.
This research strengthens the robustness and flexibility of data engineering pipelines and the underlying storage layer which form the basis of modern, scalable biomedical and health applications.
Professor Alan Fekete, Associate Professor Uwe Rehm, Professor Joseph Davis
This is an interdisciplinary field that combines computer science, data science, and public health to model, track, and forecast the spread of infectious diseases. By utilising large-scale datasets and high-performance computing, it allows researchers to simulate outbreak scenarios and evaluate public health interventions in real-time.
By advancing high-performance agent-based modeling, this work establishes a scalable computational framework and provides actionable predictive models that enable public health officials to optimise intervention strategies and mitigate the spread of infectious diseases.
Professor Mikhail Prokopenko
Complex biological and neural systems are commonly approached as dynamic networks of interacting components, utilising graph theory, dynamical systems modelling, data science and information-theoretic methods for analysis. In neuroscience for example, our research utilises statistical and machine learning techniques to analyse large-scale brain networks and extract clinically relevant patterns. This analysis focuses on both brain activity and connectivity to uncover principles of cognition, disease, and adaptive behaviour, often integrating time-series analysis with network science. Our research in biological networks focuses on genetic networks underpinning diseases, and dynamics of their mutations.
This work supports advances in understanding neurological disorders such as Alzheimer's disease and epilepsy, as well characterising normative brain function, and the broader dynamics of biological systems and networks.
Associate Professor Joseph Lizier, Associate Professor Tom (Weidong) Cai, Professor Mikhail Prokopenko, Professor Jinman Kim, Professor Zhiyong Wang
Clinical AI focuses on the development and application of artificial intelligence, machine learning, and data-driven modelling to support healthcare decision-making, patient monitoring, diagnosis, prognosis, and treatment planning. Our research brings together computational methods, clinical knowledge, biomedical data analysis, and responsible AI design to develop tools that are clinically meaningful, trustworthy, and deployable in real-world healthcare settings. Our work explores how AI can be used to analyse complex clinical data, including physiological signals, electronic health records, medical images, wearable sensor data, and patient-reported outcomes. A particular focus is placed on developing intelligent systems that can identify clinically relevant patterns, support early detection of deterioration, personalise care pathways, and assist clinicians at the bedside.
This research also considers the practical challenges of clinical translation, including interpretability, data quality, bias, privacy, safety, workflow integration, and validation in real clinical environments.
This work supports the development of AI-enabled healthcare systems that can improve diagnosis, monitoring, and decision support across a range of clinical domains. It has the potential to enhance patient outcomes, reduce clinical burden, support earlier intervention, and enable more personalised and efficient models of care.
By combining advanced AI methods with clinical expertise, this research contributes to the responsible translation of artificial intelligence into healthcare practice, with particular relevance to hospitals, community care, remote monitoring, and digitally enabled health systems. For example, our work on pain management is applying Clinical AI to support more objective assessment, prediction, and personalised management of acute and chronic pain, helping clinicians move beyond subjective pain scores towards more timely and data-informed care.
Professor Albert Zomaya, Professor Jinman Kim, Professor Zhiyong Wang
Advances in sensing technologies, computational methods, and digital infrastructure have driven unprecedented growth in the size and complexity of modern datasets. Such large-scale graph data arises in critical domains, including systems biology such as protein interaction networks, gene regulatory networks, biochemical pathways and brain networks. A fundamental challenge is that existing network analysis and visualisation techniques often fail to scale to these data sizes while preserving important structural characteristics.
Visual analytics integrates data analysis with visualisation to support human-centred reasoning over large, complex datasets. More specifically, visualisation provides a powerful computational abstraction by transforming high-dimensional relational data into geometric representations that exploit human perceptual and cognitive capabilities. However, conventional visualisation approaches for large and complex networks suffer from severe scalability limitations, visual clutter, and a loss of structural fidelity. This research addresses these challenges by developing novel visualisation methods that preserve essential network structure, enable domain experts to discover new knowledge, and remain computationally efficient at scale.
By enabling accurate and scalable visualisation of complex networks, this research will allow domain experts to perceive and interpret underlying ground-truth structures with greater confidence. In systems biology and related fields, these advances will support more reliable analysis of large-scale complex biological networks and facilitate informed decision-making in high-stakes applications. More broadly, the resulting methods will provide a foundation for the next generation of trustworthy, scalable visual analytics tools required by both academic researchers and industry practitioners.
Emeritus Professor Peter Eades, Professor Seokhee Hong
