Bringing Healthcare Data to Life

Our aim is to make rich and complex histories of population healthcare data literally visible for digital exploration, analysis, comparison, decision making, and presentation. This means developing new visual analytics, visualisation methodologies and open source software which can be applied to any UK-based collection of public healthcare data: clinical commissioning groups, NHS data, Electronic Health Records (EHR) data, and geographical information. It incorporates a multi-disciplinary team combining expertise in health research, statistics, software engineering, visual analytics, visualisation, and industry.

We live in a time of unprecedented data collection and storage driven by the low cost of storage technologies and rapid increases in computing power. However, our ability to collect and store data vastly exceeds our ability to derive useful information and knowledge from it. This is the vital role that data visualisation and visual analytics play. Visualisation and visual analytics exploit the most powerful (the fastest and most informative) human sense by far, the visual system, in order to understand and derive new and fresh insight into what otherwise might be inaccessible data. Simply put, data is much more useful when we can see it.

Collective healthcare data archive and analysis are cornerstones of the UK's healthy nation strategy. As such, tremendous resources and investment are given to ensure that the NHS is supported by the best healthcare data analytics possible. Core to this investment is exploiting novel research, technology, and data analytics that create a step change from the current state-of-the-art and informs our understanding how to support the NHS in the most efficient ways. A vital component of the strategy to support healthcare data and analysis is through major concerted efforts in big data collection, e.g., Public Health England, the SAIL Databank in South Wales, SAHSU, and InfoBase Cymru, to name just a few.

The data these initiatives collect and archive is massive and constantly growing. The SAIL database alone stores over 20 years worth of healthcare data on more five million anonymised patients resulting in billions of electronic health records (EHRs). While data collection process is a critical aspect towards improving healthcare, additional expertise and cross-disciplinary study are vital to extract useful information, understanding, and making informed decisions from the data.

This project will develop free, open source, novel visual analytics and visualisation software to study the UK's rapidly increasing volume of big healthcare data. By opening up and exposing the healthcare data analysts to visual and graphical depictions of the enormous collection of the UK's healthcare data including EHR data, a whole new digital world of health understanding, exploration, analysis, comparison and engagement is possible. This healthcare adventure coincides with a rapid rise in interest to improve the collective health of the UK.

New and accessible ways of observing UK-centric healthcare data will enable new insight and advance scholarly and public understanding of the evolution of the UK's collective health. Such understanding grows ever-more crucial with the rise in ageing population and increasing pressure on the NHS to take care of the UK's collective health and well-being.

Healthcare data analysts, data science researchers, statisticians, healthcare practitioners, visualisation researchers, students and healthcare data industry professionals will benefit. We will stimulate new work in Computer Science including visual analytics and visualisation, population healthcare data analytics, related health data and science disciplines. The impact potential is difficult to overstate. There are long-term benefits to institutions such as the NHS Public Health England, Public Health Wales, Health and Care Research Wales-all the organisations that collect and analyse healthcare data.

This project idea is interesting to many academic and other communities, and indefinitely extensible. Educators in HE health care data benefit because we will build a free visual analytics and visualisation tool that can be used in classrooms for teaching purposes including exploration, analysis, presentation, and storytelling of EHR data. The RA funded by this research project will benefit tremendously by working with leading experts in the fields of visual analytics, visualisation, population healthcare analysis and statistics. Many companies and universities are eager to hire personnel with expertise in visualisation and healthcare analytics-two core data science subjects. We have identified a number of them including We Predict Ltd., Health Economics Outcomes Research (HEOR), Digital Health Labs, and Connected Health Cities, to name a few. We have already been discussing the potential impact with these four companies. Currently, there is a severe shortage in supply for qualified experts in these fields where demand far outweighs the supply. Digital Healthcare companies may use our research prototype as the basis of an innovative commercial product. They may also use the research outcomes as the basis of future Innovate UK funding. The Medical School and the Department of Computer Science at Swansea University will benefit. Visualisation work combined with healthcare provides attractive visual material for public open days and for student recruitment events. Such projects are also interesting to undergraduate, Masters, and PhD students searching for rewarding projects. Programmers and developers will benefit from our years of development effort and expertise. They will be able to build on our open source prototype rather than having to start from scratch.

Researchers and software developers will benefit from years of application development through the release of programming source code. The free, open source tool will form the basis of an online community where public healthcare data analysis and associated expertise are shared in an open environment using existing networking infrastructure. Online user documentation and demonstration videos will be updated and released in accordance with both the release of feature updates and before workshops. The software will also be published in the Journal of Computer Graphics Techniques (JCGT). JCGT serves to bridge the gap between computer graphics researchers, and computer graphics professionals by providing a forum for practical ideas and techniques that solve real problems by publishing detailed software implementations. The software will be offered to our academic and corporate partners for further extension and improvement or the basis of a commercial product. We are already engaged in discussions with our Industrial Advisory Board concerning potential commercialisation of the result.