Image: © ipopba/Stock.adobe.com
Prof Raymond Bond believes human health and wellbeing should be central in AI development.
“I have a lot to learn,” Prof Raymond Bond says when asked about promoting his work. This understanding that there is always more to learn and enthusiasm for continuous learning has no doubt contributed to Bond’s success in his field.
Bond is a professor in Human Computer Systems at Ulster University. His PhD was in computerised electrocardiology. “This involved processing and visualising ECG data,” he explains. “The ECG is one of the most commonly used diagnostic tests to help detect heart-related problems such as arrythmias and heart attacks.”
During the PhD, Bond developed tools to visualise ECG data, including a computer simulator to demonstrate what can happen if electrodes are placed incorrectly.
“To be honest, the PhD helped highlight just how little I knew and know, and it provided great opportunities to meet international researchers from many different backgrounds, including research leaders in the medical and engineering fields. It was a humbling experience to say the least,” Bond says.
After his PhD, he worked as a research assistant and then a teaching fellow before taking up a lectureship in computer science and becoming a senior lecturer in data analytics at the Cognitive Analytics Research Laboratory (CARL) at Ulster. His interest in optimising the interactions between people and technology led to the professorship in human computer systems.
“I think we need to be humanity-centred when designing and deploying AI technologies,” he says.
“Perhaps we need to remind ourselves of what it means to preserve human values, wellbeing and happiness, and repeatedly ask ourselves whether an AI technology could negatively affect our wellbeing and happiness in both the short term and long term.
“Of course, it might be challenging to predict the long-term effects of AI technologies, but perhaps we can at least try.”
Tell us about your current research.
I am currently working on a number of projects about AI ethics, including studying the role of AI in radiography with regards to the effects of AI support in clinical decision-making. For example, how can we mitigate automation bias? Automation bias is where a person may repeatedly over trust and naively accept AI recommendations. Could automation bias result in the loss of competency in a skill such as reading x-rays or ECGs? We also have a project looking at the quality of digital health apps (using data science) as well as other projects in digital mental health and AI in cardiology. For example, one of our papers uses data analytics to show that the user ratings and the number of downloads of a health app does not correlate with the actual quality of a health app.
I am also working with others on the usability engineering of automated external defibrillators. The usability of this medical device is critical given that users (members of the public/lay rescuers) may be interacting with this user interface for the first time in a potentially life-saving scenario.
In another project, we are studying the role of biodynamic indoor lighting to enhance the wellbeing of people living with dementia. Light is important to our circadian rhythms and for managing our sleep patterns.
We also recently got awarded funding by the UK Engineering and Physical Sciences Research Council (EPRSC) for a large centre for doctoral training in digital health technologies which will provide great opportunities for new PhD researchers at Ulster University and University College London.
In your opinion, why is your research important?
I think our digital health research is important because having and promoting good health and wellbeing is obviously core to society, and we need more research to identify the positive and negative effects that digital technologies can have on our health and wellbeing.
For example, high-quality digital health interventions and apps could help prevent disease as well as helping people manage their conditions (eg adhering and engaging with treatment plans etc).
Health apps and wearables can also be used to collect health and wellbeing data ‘in the real world’, eg symptom data can be collected as repeated measures which can then be shared with healthcare staff to potentially enhance their clinical decision-making by having higher resolution personal health data. Of course, time-series symptom data can also be analysed in real time by AI algorithms to perhaps act as early warning systems and to identify key patterns/trends.
Raymond Bond
I have a particular interest in identifying methods for optimising the collaboration between humans and AI. And we need to discover and invent ways to ensure that AI is helping and not hindering clinical competence and clinical decisions. We need to create ways to calibrate how much humans should trust AI and perhaps even how much AI should trust humans (who knows). We have created an e-learning course – Introduction to AI for healthcare professionals – which may also help with promoting AI literacy amongst healthcare staff.
With regards to impact, I am hoping to see research-informed product innovation in the health and medical sectors including digital mental health innovations via collaborations with organisations such as Canary Speech, Inspire Wellbeing, Action Mental Health and Pneuma Healthcare, as well as a new biodynamic lighting device with a company called SkyJoy. Hopefully there will be new discoveries/insights to inform the design and testing of automated external defibrillators at Stryker (HeartSine). I also hope to see impact with PulseAI, with regards to the deployment of new AI algorithms for analysing electrocardiograms.
What inspired you to become a researcher?
Not sure what exactly inspired me, but I think it was the opportunity for lifelong learning. I like to learn; however, I am not necessarily great at it. I have realised that the best way to learn (for me at least) is by teaching and doing research. I find that teaching helps me structure information in a way that allows me to learn the core concepts in the right order. Research is also great because you get to mentor, present work and write. Presenting and writing does seem to help me ‘think’ about problems more deeply. The right combination of words can really help alter my perspective on a topic or set of ideas. I also naturally enjoy discussing and generating ideas.
I am lucky enough to be in a field that is broad which allows me to work in many application areas in healthcare. I really like working across disciplines and working in teams (again, you get to learn).
Science for me seems to be about questions and answers. And analysing datasets to find those answers is inspiring. I also feel inspired when I use data science/machine learning to discover new knowledge about a topic. There is little better than discovering and inventing.
What are some of the biggest challenges or misconceptions you face as a researcher in your field?
Typical challenges include planning my time to set aside hours to write and/or lead research initiatives, for example, to seek further funding. I would also love to organise my diary better so I could write more perspective or position papers on key research trends/challenges in digital healthcare (again, ‘writing is thinking’, for me at least). I must say that when I first came across the Eisenhower matrix, it helped me think more about the tasks that are really important but not urgent. I need to avoid neglecting these ‘important but not urgent’ tasks.
Do you think public engagement with science and data has changed in recent years?
I have not done any analysis to see if there has been any significant changes in the public engagement with science. Perhaps given the pandemic, more citizens might have engaged more with topics such as exponential growth, statistics, graphs and modelling, which could have potentially contributed to improving statistical literacy. Readers might be interested in a paper I co-authored titled Why Pandemics and Climate Change Are Hard to Understand and Make Decision-Making Difficult.
How do you encourage engagement with your own work?
I try to work with stakeholders and companies when doing research. This is mainly to work towards having a real-world impact on products and stakeholders. Collaborating with companies reminds me of the power of the Ikea effect which suggests that we can perhaps attribute a lot of value to the things that we build ourselves using our own time.
I try to encourage press releases about our work; however, I would like to engage more with standards and policymakers.
Networking and giving talks at different events has also helped promote our work and I have tried to integrate my research into my teaching. For example, it has been really great to see MSc students go on to publish their work in research forums. I was also involved in creating a research-teaching nexus toolkit to help promote the integration of research into teaching across the university.
Find out how emerging tech trends are transforming tomorrow with our new podcast, Future Human: The Series. Listen now on Spotify, on Apple or wherever you get your podcasts.
