The NIHR Research Design Service, Health Data Research UK (HDR UK), the MRC Regulatory Support Centre and NHS Digital are pleased to offer the opportunity to attend the sixth in a series of virtual researcher roadshows.
This free to attend event is aimed at researchers and others working with health and social care datasets who would like to increase their understanding of roles of health data scientists and value of health data science.
The event will bring together health data scientists from the NHS, academia, and industry and provide insights and perspectives on the important work they do. It will include presentations on different roles and careers in health data science and different methods and approaches used by health data scientists.
Agenda
| Event timings |
Agenda item |
Content of session What will participants get from it |
Names of Speakers |
| 10.00 - 10.05 |
Welcome and introductions |
Introduction to event, housekeeping, run through agenda and poll to find out more about the audience |
Chair – Martin Williams |
|
10.05 - 10.50
45 minutes total
(15 minutes per speaker and 15 minutes for questions)
|
Methods of Health Data Scientists
|
Methods Health Data Scientists use to solve problems/ address questions (noting the context of other methods)
|
Fionna Chalmers (BHF Data Science Centre)
Watjana Lilaonitkul (UCL)
|
|
10.50 – 11.55
|
Introduction to Health data Science Panel discussion
|
Each speaker gives a short (3 min) introduction to themselves and their career pathway before panel discussion followed by questions from the audience.
Moderating questions will be Alex Bailey (MRC UKRI) and Martin Williams (NIHR)
|
Andrew Boyd (University of Bristol, UK Longitudinal Linkage Collaboration – Chair)
Palwasha Khan (LSHTM)
Richard Savage (Pinpoint)
Zoe Turner (NHS-R)
Emma Crawford (MSD)
|
| 11.55 - 12.00 |
Event close |
Poll to gather feedback.
What happens next, follow up contact, future events |
Martin Williams (National institute of Health Research (NIHR) |
Further details on talks
Fionna Chalmers - BHF Data Science Centre
Health data science is an exciting and constantly evolving area, which means that data scientists in the field must be committed to continuously learning new methods and techniques.
Fionna, an Early Career Health Data Scientist working within the BHF Data Science Centre, will talk about a recent piece of work she undertook and her approach to staying curious and learning new data skills.
She will discuss the tools she uses to solve a variety of problems and how the BHF DSC use a team science approach to improve health with open source, reproducible, research.
Waty Lilaonitkul - UCL
AI applications in health and healthcare have demonstrated human-level performance across various medically related tasks but have yet to show real-world impact in practice.
In this talk, Dr Waty Lilaonitkul will discuss her work on building human-in-the-loop AI pipelines in collaboration with healthcare practitioners.
This collaboration aims to create an agile design-train-test loop in technology development via new algorithms to concurrently optimise performance and robustness in data-starved environments. By iteratively combining the best of human and machine intelligence to force more efficient learning paths, machines are able to actively learn better features from fewer training examples at significantly faster rates.
As case-in-point, she will discuss (i) PKPDAI.com, an opensource AI-backed system that automatically 'machine-reads' publications to continuously extract and grow a pharmacokinetic database across all drugs to support AI research in drug development, and (ii) AI4IRDs, a pipeline for multimodal research in rare diseases, which has yielded 3 patent-applications.
Panel discussion
The Researcher Roadshow panel, chaired by Andy Boyd (University of Bristol, UK Longitudinal Linkage Collaboration), will introduce themselves and their varied routes into health data science. The panel will discuss pathways into health data science careers and there will be a large amount of time in this session dedicated to audience questions.
Watch a recording of the webinar