Where are we with AI in healthcare? Faculty’s Alistair Stuart and Livingbridge’s Sanjay Panchal explore emerging trends

Livingbridge Partner and Head of Healthcare investment, Sanjay Panchal, sat down with Faculty Director, Healthcare & Life Sciences Alistair Stuart to explore the latest developments and trends at the intersection of healthcare and AI. Livingbridge currently has three investments playing into the wider adoption of technology in Healthcare. Faculty, founded in 2014, is a leader in Applied AI. Faculty help clients navigate the AI transition. From strategy and roadmapping to the design, build and deployment of bespoke AI systems that deliver productivity and profitability.

As health systems struggle to bridge the gap between a shortage of supply and ever increasing demand there is a strong imperative to drive adoption and shake up the status quo. This combined with the enormous amounts of data being generated in healthcare, and our increasing ability to measure more make it a perfect environment to leverage the full power of AI solutions by improving outcomes, driving productivity and increasing confidence in health systems more broadly.

The challenges to this are largely a question of appropriate change and risk management, cultural buy-in and creating the time for clinicians and carers to learn and embed new technology solutions in their ways of working; and of course…funding.

Sanjay said he is already seeing a surge in investment interest and deployment of AI in medical diagnostics and operational decision-support tools. Livingbridge’ portfolio company Everlight Radiology, a teleradiology services provider, is using AI tools to provide additional reassurance to clinicians in their decisions and they expect this to ultimately evolve into better triaging and case allocation as confidence in the systems available increases. Another portfolio company T-Pro, is using machine learning to deliver a faster and more accurate documentation process for clinicians, saving hours of time per week and allowing clinicians to spend more time with patients.”

Another successful example is in community remote monitoring where AI tools are now being used to predict changes in an individual’s condition and flag to clinicians where interventions may be needed sooner to prevent further deterioration or the need for hospitalisation. This is a real win-win as it results in better patient outcomes, more cost-effective resource management and ultimately lower care costs.

Sanjay said: “I see technology as having three broad types of impact on healthcare. The first is improving efficiency of care through removing routine administrative add tasks - allowing more people to be cared for with the same resources and enabling clinicians to spend more time with patients.

“The second is improving the quality of care in the broadest sense. It doesn’t just mean better and faster clinical outcomes, but the quality of the patient experience and patient satisfaction.

“The third is delivering care more consistently across a healthcare system, and reducing variance in outcomes.”

Here, companies like Faculty are already making a difference. Alistair said: "Healthcare-related AI technology is currently going through a period of unprecedented development, but we have still only scratched the surface in terms of the right use cases, and how today’s models can be deployed within healthcare. There's a massive opportunity to make health services more efficient overall.

"For example, many patients currently occupying hospital beds no longer meet the criteria for hospitalisation - they could be discharged if the health system could process the discharge (think pharmacy needs, equipment, transport of social care needs that must be in place prior to discharge). Keeping people in a bed for as short a period of time as possible leads to demonstrably better patient outcomes. Therefore being able to predict and plan the discharge needs of a patient might be on their day of admission, then timely discharge becomes more possible, and valuable resources are released across the healthcare system."

However, both Alistair and Sanjay caution that AI adoption and system-wide transformation will certainly move slower than the pace of technological development, in part due to perceptions around AI and medical data privacy.


Sanjay says: “There are debates going on around how protected people's data should be and in what form, and the extent to which anonymised healthcare data should be open-access. Healthtech companies need big datasets, but current regimes prevent companies from making the most of that data because it has to sit in siloes, so this is an journey we need to go on.

"We need to get comfortable sharing our healthcare data for the greater good, and at the same time feeling safe that it’s not being exploited and that it’s not identifiable.”

Alistair highlighted Faculty's recent work with synthetic data sets as part of an important trend helping companies navigate this data issue in the short term.

He says: "You can take a dataset within a provider, look at its characteristics and make synthetic data that looks similar to that population. This means you've created non-patient data, and you can use that for your work. This is a technology solution that’s starting to come in.

"There are limitations to synthetic datasets, but the advent of GenAI is interesting here as it unlocks new avenues for these synthetic datasets. You can use large language models (LLMs) and GenAI to enhance the quality of synthetic data. That's improving all the time and it's another way to overcome these privacy concerns.”

Faculty worked with Genomics England, the historic UK life-science program analysing 100,000 sequenced genomes to help in disease prevention and treatment. The project saw Faculty put disparate data sets together for the first time to build a modelling framework that could help predict overall survival rates for oncology patients.

"That demonstrated it was possible," Alistair says. "The sector is rightly conservative about new technology, but there is also significant potential in the amount of unstructured data available in healthcare systems, from EMRs through to scientific literature."

Looking longer-term, Sanjay stressed that “we need to get to a 2.0 of healthcare technology systems”. Both Alistair and Sanjay believe this will involve moving from using single data sets in silos to combining data sets from various sources to develop richer insights. Today for example an individual’s diagnostic, genomic and health record information will largely be generated and used in isolation, or aggregated in a laborious and manual fashion. We will move to a world where the ability to combine these and derive useful research and clinical conclusions will help improve pharmaceutical R&D and allow for better and more personalised care pathways.

Alistair says: “It will take time to work out where the value lies and how to use all these new tools in the best way within the healthcare system. These things need to be dealt with caution so that we don't expose patients to risk, and so that tools are proven before they reach the mainstream.

"It's very exciting, but there's still a lot of work to do.”