LIDA Annual review – introduction and overview by Professor Alejandro Frangi
Research with impact – benefitting populations and changing our world
I am pleased to introduce our world-leading, award-winning, and innovative work in complex modelling, statistical data science, machine learning and AI in our second online LIDA Annual Showcase. As leaders, not followers, in data analytics and AI, we continue seeking to drive change through data across multiple disciplines, working with academics, industry, and organisations around the world to benefit populations on a global scale.
International symposiums conducted online over the past year meant collaborations could continue despite the impacts of Covid-19, meaning our important work encompassing health, societies, and environment has continued despite challenging circumstances.
Leif Denby’s work at LIDA is part of an international collaborative research project examining the effect of cloud organisation on Earth’s climate . It uses the world’s first unsupervised neural network model to autonomously discover cloud organisation regimes in satellite images, with important implications for weather prediction and climate modelling. A five-year European Research Council-funded project led by Prof. Andy Hooper and Dr. Matt Gaddes uses complex data modelling work, radar interferometry and an algorithm they built, LiCSAlert, to detect, measure and forecast volcanic deformation and eruptions from space.
Driving innovation in climate-smart food systems, Sam Bancroft and Prof. Andy Challinor’s work is linked to Leeds University Farm. It uses existing crop models, AI, deep learning (a subset of machine learning) and remote sensing to look at how crops can be managed to increase food security. See the full case study here.
Dr. Jianhua Wu’s national cardiovascular care and outcomes research has revealed quality of care for cardiovascular disease (CVD) patients increased during the pandemic but the number of patients seeking care decreased. The work has highlighted the need for clarity in public messaging to prevent unintended consequences of social distancing mandates. A multidisciplinary team led by Dr. Ramesh Nadarajah has used Newton Gateway modelling of NHS CVD waiting lists to examine how pandemic impacts on CVD patient care can be lessened. It has led to feasible, pragmatic solutions to bring down waiting lists and reduce deaths.
A study led by Dr. Andres Diaz-Pinto, Dr Nishant Ravikumar and me in collaboration with Prof Sven Plein and Prof Chris Gale, funded by the British Heart Foundation and the Royal Academy of Engineering, seeks to identify early indicators of cardiaovascular disease without the need for more costly and less accessible cardiac MR imaging examinations. In a transcontinental collaboration involving the UK Biobank and the NIH AREDS studies, we demonstrated that the eye can indeed be a window to the heart and allow prediction of heart attach with only a photograph of the eye fundus and minimal personal information. In a separate study led by Dr Ali Sarrami-Foroushani and me and funded by the Royal Academy fo Engineering, we undertook the most comprehensive in silico clinical trial of flow diverters to treat cerebral aneurysms to date , which could reduce costs and time spent in clinical trials from years to a few days. This work, published by Nature Communications demonstrated AI and computational modelling can bring about a transformation to regulatory science and innovation.
Prof. Richard Feltbower and Dr. Adam Glaser’s ongoing work on international childhood cancer survival involves the only UK database running continually from ages 0 to 29, the Yorkshire Specialist Register of Cancer in Children and Young People (YSRCCYP). It informs the decision-making of local clinicians and commissioners with collaborative national and international epidemiological and outcomes research. The YSRCCYP is currently examining health inequalities data, specifically ethnicity variation in cancer survival. Read the full case study here.
The Local Data Spaces project has been recognised as delivering public good and informing policy decisions, winning the ONS Research Excellence Project Award 2021. This data analytics project supports local authorities, groups, and stakeholders, using granular, secured data and research-driven analyses to ensure proactive responses to Covid-19 at the local level. Dr. Nik Lomax’s work on SPENSER (Synthetic Population Estimation and Scenario Projection Model), a set of open-source tools for population estimation and projection funded by the Alan Turing institute, provides high-resolution geographical and sub-population projections that are essential for planning and delivering services and developing urban infrastructure.
A collaborative, multi-disciplinary and multi-institution project initiated and led by Dr. Robin Lovelace, Asst. Prof. Transport Data Science here at University of Leeds, is revolutionising design, construction and strategic cycle planning in England and Wales with the Propensity to Cycle Tool used by over 35,000 transport planners, consultants, advocates, and members of the public.
The work we undertake at LIDA, as evidenced throughout this showcase, has real world impact with demonstrable value, showing data can be used across all aspects of society to directly benefit populations.