Haris Shuaib is the AI Transformation lead at the London AI Centre which is a hub of innovation in healthcare in the UK, recently receiving £16m in funding to scale up AI work in the NHS and being appointed a preferred partner for NVIDIA’s proposed £40m healthcare supercomputer. With a background in theoretical physics, Haris trained as a medical physicist prior to moving into a leadership role in AI in the NHS.
Getting AI to work in real-life healthcare.
Haris discusses the variety of healthcare AI projects the London AI centre are running and how they select which areas to prioritise. There are many challenges in getting AI integrated into clinical pathways and Haris explores many real-life examples of AI being deployed in NHS hospitals from paediatric bone age xrays to head CT scans to breast screening technologies.
The London AI centre focus on getting systematic processes in place for evaluation and focus on acceptance testing prior to deploying an algorithm into a clinical pathway. Projects usually have tripartite leadership involving clinicians, academics and industry partners which allows for rich collaboration.
Guy’s hospital also has a very helpful commercial team who help support clinicians to spin-out companies using technologies created in the hospital environment.
Clinician scientists and NHS leadership
Clinicians scientists have a unique lens through which they view the deep tech healthcare world yet they don’t traditionally take leadership roles in the NHS.
Haris discusses his time on the Topol Fellowship leadership scheme which allowed him to grow as a leader and have the confidence to amplify his voice as a clinician scientist. This led to him stepping forward into a leadership role and encouraging other clinician scientists to do the same.
As AI is adopted on a wider scale in the NHS, the challenges it brings are very technical. Thus, the skillset of clinician scientists is even more pertinent in making technically sound decisions around the deployment of AI. As clinical data science leadership grows in AI communities in healthcare, clinician scientists are well placed to drive innovation in this field.
NVIDIAs healthcare supercomputer.
NVIDIA have made a £40m commitment to build a supercomputer in Cambridge. Named ‘Cambridge-1’, it is intended for AI research in healthcare. The London AI centre are one of the key founding partners chosen to work with NVIDIA on this dedicated healthcare research. Some of the projects being explored include building an ‘image net’ of medicine and linking GP data to hospital data to help predict the next pandemic or to allow earlier cancer detection.
The big topics in healthcare AI
Haris is pragmatic about the challenges of AI and acknowledges that there are a lot of unknowns that reveal themselves as the field develops at a rapid pace. He delves into the big questions that need to be addressed in healthcare AI.
- What happens when there is discordance and conflict between the AI and doctors?
- How can we involve patients in co-designing AI technology?
- What can we do to reduce bias in AI and improve diversity in both the AI workforce and in the data used to create the algorithms?
- Why isn’t informed consent fit for purpose?
Haris’ bold takes on the future of imaging include imaging on the high street with mini-MRI machines and AI taking over the traditional role of the doctor in the new information age.
If you’re in healthcare and have an interest in the future, this is the podcast for you.