A-Z of AI in Healthcare

AI that plans, uses tools and acts in multiple steps to reach a goal.

A standard used to compare an AI model’s performance against alternatives.

A mathematical test that can be used to assess an algorithm's performance.

An exact digital replica of a real-life person, thing or process.

The moral guidelines that guide decisions about what should be done (or not).

A large AI model pretrained on broad data for many tasks.

A way to determine how well an algorithm works in a new setting.

The process of ensuring data from different sources can be used to train AI.

How understandable an AI model's decisions are to a human.

Strategies used to improve efficiency, add more meaning and challenge to jobs.

A way or organising data from different sources to clarify their relationships.

A type of AI trained on large datasets that generates text or images.

Data that is captured in different sources and formats and then joined together.

A type of AI that uses interconnected networks of neurons, like a human brain.

An error that occurs when an algorithm becomes too familiar with training data.

Using AI to estimate an individual's risk of a specific healthcare outcome.

How much data is available to train AI vs how suitable that data is for training

The format in which data is stored, analysed and processed by devices.

Understanding which groups of patients respond to drugs and in what way.

The process of teaching an algorithm how to do its job.

Two common methods for training AI algorithms in the context of healthcare.

The process of testing to see whether an algorithm works well.

Algorithms that are transparent by design vs those that are not explainable.

A field of research that aims to make AI algorithms easier to understand.

How healthcare professionals are essential to the responsible use of AI.

The ability of an algorithm to classify an image that it has not seen before