Navigation

3 new NHS-funded AI healthcare developments for 2021

Health & Education

New AI systems are set to dominate healthcare in 2021 and the following years.

Demonstrating the rising interest in the space, more than 40 new projects have joined the NHS’ AI Lab as part of its AI in Health and Care Award Programme since Sep20. The programme, announced by Matt Hancock in 2019, aims to invest a total of £140m to fund budding AI-based healthcare projects. Here are three of our favourites from the recent announcements:

1. EchoGo, Ultromics

EchoGo, developed by Ultromics, is a new cloud-based AI service used to aid in predicting heart disease, one of the UK’s leading causes of death each year.

The system analyses ultrasound scans of patients’ hearts using machine-learning algorithms – software that automatically improves itself based on experience. The system is said to provide near-instant results with a good degree of precision – something which, due to the nature of machine learning, is likely to improve still further with time.

The goal of the software is to increase working efficiencies for clinicians, but also to speed up diagnoses and improve accuracy. One advantage software has over the human brain is the sheer amount of processing power it can call upon. A human can only really process a single thought or calculation at a time, and that process will take, say, a few seconds. Computers, on the other hand, can process billions of calculations every second. For this reason, computers are often better than humans at analysing massive data sets, such as ultrasounds, making them ideal for these pattern-spotting tasks.

2. BioEP, Neuronostics

The BioEp platform aims to make the diagnosis of epilepsy faster and more accurate, as well as monitoring the effectiveness of the drugs used to treat it.

The new AI system works by analysing segments of electroencephalogram recordings – recordings of the brain’s activity – and creating a model of the brain from this data. This allows for accurate, personalised models of the brain to be built and tested upon, allowing for more specific treatment.

The platform is faster than the conventional method which involves extensive analysis of brain recordings to determine the result. With the job being done by an AI-based system, efficiency and accuracy should improve and keep on improving as the system analyses more data.

3. Mia, Kheiron

Mia is a new AI system developed to replace one of two radiologists mandatory in each breast screening process. Mammograms, the machines used to screen patients, are read by two independent radiologists in order to assess the patient accurately; Mia uses AI to detect patterns in the mammogram readings data, essentially filling the role of one of the radiologists.

The system has been tested in 40,000 mammograms, returning near-identical cancer detection as a human radiologist an impressive 80% faster. Unlike a fully AI-based system, Mia still allows for the nuanced and experience view of a human practitioner to function alongside its algorithms. The system strikes a good compromise between meeting the healthcare workforce crisis and retaining an element of human input.

From looking at all these systems it becomes clear that the aim of AI is to increase efficiency and accuracy by automating data-intensive tasks, therefore freeing up the human workforce to pursue more nuanced and creative tasks. Given the nature of machine-learning, these systems should just keep on iterating and improving the more data they collect; the future of the AI-aided workforce looks bright.

If you have an insight into AI and machine learning please do get in touch, we have plenty of clients who need some!

In September, 42 AI projects joined the NHS AI Lab… £140m is to be invested over three years to help to accelerate the testing and evaluation of AI technologies.

By Rebecca Garland on 22/12/2020