Current Research Focus: Type 1 Diabetes

In 2019 I pivoted my entire research portfolio to Type 1 Diabetes technology, a transformative shift driven by the opportunity to apply two decades of work in digital phenotyping, user modelling, and wearable sensing to one of the most data-rich and clinically pressing challenges in modern healthcare. The research now spans a multidisciplinary team built across the NHS, Diabetes UK, the Faculty of Biology, Medicine and Health, and the Manchester Diabetes Institute. Central to this effort is Melontech, a start-up I founded that is currently passing gate 3 with the UoM Innovation Factory after an initial £20,000 investment, creating a faster pipeline from research findings to clinical practice. Alongside this, I have established the Diabetes and Hypoglycaemia Technology Network at the Pankhurst Centre and assembled a Brains Trust of 80 citizen scientists living with Type 1 Diabetes, whose direct involvement shapes the direction of the network and its technology development.

Grant income has grown steadily as the work matures. Recent awards include the Wellcome Trust Translation Grant of £25,000 (2022/23), the MRC Translation Grant of £25,000 (2023/24), the combined MRC/Wellcome Trust Confidence 4 Translation grant of £65,000 (2023/24), and the Pankhurst Seedcorn Funding of £15,000 in 2024. A further EPSRC Healthcare Grant of approximately £850,000 is currently under review. Eleven journal papers in high-impact diabetes and hypoglycaemia venues have already emerged from this work, addressing hypoglycaemia prediction, continuous glucose monitoring, and behaviour change in the management of the condition.

Research Areas

My research spans several interconnected areas that have developed across a twenty-five year career, each informing the others as methods and insights migrate across application domains.

Digital phenotyping and personalisation lie at the core of the current diabetes work, but have their roots in earlier studies of web navigation, eye-tracking, and the behavioural signatures of people with disabilities and neurological conditions. The central ambition is to build adaptive user models for people operating in extreme conditions — whether those conditions are defined by a rare metabolic disease, Parkinson’s Disease, or severe visual impairment — and to use those models to deliver personalised, timely interventions. The approach draws on continuous passive sensing from smartphones and wearables, longitudinal data collection in naturalistic settings, and machine learning methods capable of detecting subtle shifts in behaviour that precede clinical events.

Web accessibility represents more than twenty years of sustained contribution to a field I have been instrumental in establishing as a rigorous scientific discipline. This has happened through conference creation (the ACM W4A Conference, now in its twentieth year), journal editing, active participation in international standards through the W3C, and a consistent effort to raise the empirical standards of the field and make data, tools, and methods openly available. The outcomes of this work have found their way into the DNA of the web: into the Google Chrome browser, into the Eclipse ACTF framework used by accessibility tool developers worldwide, and into W3C standards including UAAG 2.0, which governs how screen readers interact with dynamically generated web content.

Ambient intelligence and user modelling research addresses the challenge of predicting and influencing behaviour in instrumented environments, from smart homes that infer activity patterns from sensor streams to clinical settings where passive data can reveal the progression of chronic disease. Work in this area has produced novel methods for activity prediction, intrinsic motivation recognition, and social sensing using smartphone data, with applications spanning Parkinson’s Disease, COVID-19 behavioural compliance, and bipolar disorder.

Health technology applications tie these threads together. For Parkinson’s Disease, my team collected the largest longitudinal dataset in the field, comprising 300 million data points across ten participants generating fifty million data points per person per month from twenty-nine data sources and sensors, alongside clinical and self-reported measures. For autism, eye-tracking methods developed in accessibility research have been adapted into detection pipelines based on web-interaction scan paths. For COVID-19, digital phenotype methods were repurposed to understand compliance with public health policy.

Research Collaboration

The research has always been built on partnership, both industrial and academic. Current and recent industry collaborators include NHS trusts, Diabetes UK, Novonordisk, Eli Lilly, BBC Research and Development, Google, Intel, IBM Research at both the T.J. Watson and Tokyo centres, Philips Healthcare, Ove Arup Partners, Telefónica/CTIC, and DoCoMo. International academic collaborations span the University of Oulu Medical Faculty and the Academy of Finland ICT 2023 programme, Carnegie Mellon University, and the University of the Basque Country.

The Interaction Analysis and Modelling Laboratory, which I established and equipped entirely through competitive bids, underpins much of this collaborative activity. The lab brings together human factors specialists, psychologists, and computer scientists and is equipped with a two-room usability suite, three eye-trackers, three galvanic skin response monitors, and full audio-visual recording systems. Its facilities are used by external organisations including the BBC, Thomson Reuters, and NICE. Embryonic IAM laboratories have grown up at the Middle Eastern Technical University and the University of Udine, forming an informal European network with shared research interests and active researcher exchange.

Postgraduate Researcher Supervision and Development

Training the next generation of researchers is something I regard as one of the most significant and enduring aspects of my contribution to the field. I have supervised 26 PhD researchers to successful completion and currently supervise a further 13, making my group one of the larger active HCI and digital health research groups in the UK. Every one of the 26 graduates has gone on to a successful career in academic or industrial research; they include the current Head of Health Data Science at the MHRA-CPRD, academic staff at universities in the UK and internationally, and senior researchers in major technology companies.

Of the 37 PhD researchers I have supervised in total, 14 brought external funding to the department, generating a career total of £421,000 in PGR income, with £171,000 falling within the last five years and a further £277,200 currently on programme from researchers who would not have come to Manchester without my specific projects.

Current students are working across the full range of the group’s interests. Alex Hambley is completing work on optimised web accessibility evaluation; Nicole Lubasinski and Daniel Gasca García are working on nutrition analytics and blood glucose prediction in Type 1 Diabetes; Naziha Rida Mohamed, Nuo Cheng, Joshua McDonagh, Ziyang Zhou, and Ahmad El Cheikh Ammar are pursuing projects in digital phenotyping and health technology; Hatim and Ashwaq Alsayahani, Vee Kongdee, Ahmad Bilal, and Mohammed Basheikh are working on topics in accessibility, sensing, and user modelling.

Beyond direct supervision, I served as Director of Postgraduate Research for Computer Science, overseeing 240 PGR students. During that period I initiated substantial changes to EDIA support, seeing the department’s figures surpass the AdvanceHE averages for both Computer Science and Engineering; drove PRES satisfaction scores above the University average; increased four-year submission rates from 38 per cent at the start of my term to 76 per cent at its conclusion; and raised first-time pass rates from 54 per cent to 89 per cent. I have also served on the Faculty PGR Awards Committee and the Faculty PGR Committee, and continue to line manage seven members of staff from early probation through to full professor.

The six post-doctoral research associates for whom I have been Principal Investigator have all gone on to successful careers in academic and industrial research. Supporting researchers at every career stage, and ensuring that the environment of the group is one where intellectual ambition and rigorous method are both valued, remains a central priority.

Research Impact

The metrics for my full career reflect both the volume and the sustained influence of the work. I have published over 160 papers, accumulated 6,014 citations, and hold an h-index of 44 and an i10-index of 129. Google Scholar and Microsoft Research both rank me in the top five per cent internationally for HCI and the World Wide Web. Scholarometer places me 98th internationally for HCI using its career and domain normalised hs-index, while ACM Bibliometrics record 47,131 cumulative downloads across 126 papers, averaging 524 downloads per article.

ACM President Alexander Wolf described the work as having “singular impacts on the vital field of computing” with “achievements that have had a significant influence on the social, economic and cultural areas of daily lives all over the world.” Sir Tim Berners-Lee commended it as “instrumental to the development of the Web.” Awards recognising the research include the ACM Doug Engelbart Prize (2000), the ADDW IBM Research Prize (2005), the Microsoft Web Accessibility Judges Award at W4A (2008), and the Best Paper Award at W4A (2010).