‘A window into the heart of Artificial Intelligence’ was submitted by BHF-funded researcher Prof Michelle Williams, of the University of Edinburgh and BHF Data Science Centre. This bright neon representation of a heart has actually been created by the AI image generator Dall.e, from the sentence “An image of the heart from a computed tomography scan, digital art”. Whilst able to create a stunning image, the AI has produced a few mistakes in the anatomy and structure of this heart, most noticeably in the branching pattern of the blood vessels.

Prof Williams is looking to see whether healthcare professionals can spot the differences between AI-generated images and real images taken from heart scans. She is also focusing on how to use AI to improve the accuracy and detail of information from heart scans, like the computed tomography (CT) scan imitated in this image.

Prof Michelle Williams said: “AI has huge potential to improve the treatment and care of heart and circulatory disease patients, including by enhancing the diagnosis and management of these diseases with heart scanning. However, it is important to remember that, at this stage of development, it can also introduce mistakes and biases.

“I’m very pleased to make the shortlist for the BHF’s image competition, and for the opportunity to draw attention to these issues.”