Internet Explorer is no longer supported by Microsoft. To browse the NIHR site please use a modern, secure browser like Google Chrome, Mozilla Firefox, or Microsoft Edge.

Artifiicial Intelligence in Health and Care Award - AI definition

Contents

Published: 30 October 2020

Version: 1.0

Print this document

NIHR currently has three routes for researchers and industry partners seeking funding for health research involving artificial intelligence (AI). This document outlines how the NIHR Artificial Intelligence Health and Care Award defines AI.

How does the AI Award define AI? 

There is no single, universally agreed definition of AI, nor indeed of ‘intelligence’. Broadly speaking, intelligence can be defined as ‘problem-solving’, and ‘an intelligent system’ as one which takes the best possible action in a given situation. We are using the definitions described by the AHSN network.

The ‘A’ of AI generally refers to one of the following:

Artificial (intelligence) 

This makes it possible for ‘machines’ to learn from new experiences, adjust outputs and perform human-like tasks. It can be thought of as the simulation of human intelligence and could include voice and visual recognition systems.

Augmented (intelligence)

These are outputs that complement human intelligence, emphasising AI’s supplementary role. Examples include tools that support radiologists in reviewing large numbers of scans, or that support financial advisors to better understand clients’ current and potential future financial needs.

Ambient (intelligence) 

The application of several technologies (including Artificial or Augmented Intelligence, but also sensor networks, user interfaces, home automation systems, etc) to create proactive ‘smart’ environments.

Complexity scale in AI

There is a significant amount of effort being devoted in the research space to map machine learning and AI, but it has been challenging to categorise them according to their ‘intelligence’. Thus far, attempts at categorisation have been limited to looking at their generic ability to solve new problems, and at the speed with which they adapt to these problems. A more straightforward way of understanding AI is to classify AI systems by their complexity. This has been described here. We encourage AI solutions of all complexity, however higher complexity AI solutions may be more valued in the assessment.

What is not artificial intelligence?

Due to the complexity scale in AI, it can be hard to define what is not artificial intelligence. When considering the AI solution, we will consider what problem you are trying to solve, and whether artificial intelligence is the right solution. You should be able to explain why you are choosing AI, what additional ‘intelligence’ do you need, and why AI is the solution. You should consider:The problem you’re trying to solve is associated with a large quantity of data which an AI model could learn from 

  • Analysis of that data would be on a scale so large and repetitive that humans would struggle to carry out it effectively
  • You could test the outputs of a model for accuracy against empirical evidence
  • Model outputs would lead to problem-solving in the real world
  • The data in question is available – even if disguised or buried – and can be used ethically and safely.

If you can’t satisfy these points, a simpler solution may be more appropriate. More information is available in NHSX’s Buyers Guide to AI in Health and Care.