Research

Modelling and Simulation

…will expand research from the Virtual Physiological Human (VPH) initiative to lead pan-European developments in modelling and simulation. A rich assortment of complementary modelling approaches is already available, including 3D (e.g. Navier-Stokes solvers, Lattice Boltzmann approaches), 1D and 0D models of fluid mechanics for cardiovascular and respiratory applications, as well as Finite Element Analysis, Growth and Remodelling frameworks, and Agent-Based Models of structural mechanics for cardiovascular, musculoskeletal and oncological applications.Read more

Multiscale approaches which facilitate integration at various scales are required to represent the full extent of biological complexity, addressing problems which involve strongly coupled physics (e.g. heart valve mechanics) and biological processes that evolve with strong dependence on the local mechanical environment (e.g. bone remodelling). While High Performance Computing provides solutions for significant computational loads produced by such applications, for clinical workflows that must operate inside a time-critical pathway, methods like Reduced Order Modelling will be advanced. Sano modelling pipelines will address physiological simulations from a multi-morbidity perspective through integration with healthcare-oriented data sources and analytical technologies.
Modelling and Simulation
High Performance Computing

High Performance Computing

…will address fundamental computer science challenges related to Modelling and Simulation and In Silico Techniques. The computational and data processing needs of the Centre will require pushing the boundaries of current state-of-the-art HPC and cloud infrastructures. This includes research to align the capabilities of modern computing environments with the computational needs of tools, workflows and institutional systems to deliver patient-specific in silico care on clinically-viable timescales.Read more

New developments in computer hardware, programming models and IT services will influence the way in which traditional computational models are developed, deployed and executed. Delivering state-of-the-art in silico services will also call for exascale computing resources. Consequently, the Centre will become a facilitator of change within EU HPC initiatives, including PRACE, EuroHPC and their successors. Close integration with Data Science will match computational resources to the processing needs of machine learning – one of the most demanding and rapidly developing areas of modern computer science.

The Centre will also work to develop new international healthcare standards, promoting sustainable software development and addressing security/privacy issues – particularly important for medical applications where the strictest regulatory requirements apply.

Data Science

…will address challenges associated with processing healthcare data, including medical images. Applying tools developed by the Centre both for entire populations and for dense data sources describing individual patients will require further advancement of the state of the art in Big Data processing and close integration with research in Healthcare Informatics.Read more

This will enable monitoring data to directly feed into diagnostic and treatment algorithms, and to provide personal health forecasting. Machine learning techniques will be applied to assist with data mining, identification of missing data in sparse clinical records and optimisation of analytical algorithms. This activity will develop data analytics tools to aggregate diverse data sources and automate reasoning logic, providing novel diagnostic insights for patient populations with complex clinical histories and significant comorbidities.
Data Science
In Silico Techniques

In Silico Techniques

…will develop the methodology to extend in silico techniques to various aspects of healthcare. Sano will integrate individual in silico patient stratification and treatment planning across many clinical domains, allowing the influence of multi-morbidity to be assessed within the in silico treatment pathway. Our ambition is also to advance the emerging area of in silico clinical trials, by applying modelling and simulation to evaluation of devices, drugs and interventions over enormous trial populations, with HPC infrastructures being leveraged to bring down the required time and cost.Read more

Significant research challenges related to missing or inaccurate data will require integration with the Data Science and Healthcare Informatics areas for input data distribution and accuracy, and with the Modelling and Simulation area for output uncertainty determination. In silico methods will also be used in personalised health forecasting, ensuring real-time availability of up-to-date personalized patient prognostics based on constantly refreshed personal data, shifting population’s health philosophy from reactive to proactive focus on optimised living. This change is likely to be the single most significant alteration in social behaviour ever to have followed a technological breakthrough.

Healthcare Informatics

…will address challenges associated with integration of next-generation scientific tools, especially computerized decision support system, with healthcare control systems. The required techniques include theoretical and practical integration methodologies to benefit from systems in widespread clinical use (PACS, Clinical Management Systems, etc.), as well as improving existing procedures to maximise the availability of anonymous data while maintaining its integrity. Read more

Along with developments in Data Science, this research area will advance missing-data correction algorithms, permitting In Silico Techniques to quantify uncertainties in model predictions, both individually and across patient populations. Together with Data Science and Algorithmic Decision Science, this area will be central to the design of clinical support tools that represent the Centre’s ultimate technological output. All this will be done in close collaboration with clinical partners and healthcare administration communities to ensure integrated system optimisation.
Healthcare Informatics
Algorithmic Decision Science

Algorithmic Decision Science

…will integrate the five preceding areas into Decision Science: a novel international academic discipline. By introducing a formalised, rigorous structure for the development of fine-grained, evidence-based rules, operating on up-to-date and thoroughly validated input data, it will supplant expert systems as the basis for machine-enhanced decision support. Read more

Able to integrate the full breadth of domain knowledge, this new science will enhance the efficacy of treatment by accounting for the effect of co-morbidities on treatment outcomes. More significantly, it will enable novel treatment strategies to be developed for multi-morbidity patient groups, driven by quantitative assessment of the impact of each contributing disease, of their interactions, and of various treatment options. Such an approach will transform the healthcare paradigm to promote the value of modelling and analysis of healthcare data beyond the boundaries imposed by the traditional organisational models of clinical specialties.