Conventionally AI is used to automate (typically post-acquisition) single image analysis tasks that a human expert would otherwise perform such as anatomy delineation, measurement and image alignment.

AI-enabled biomedical imaging refers to an emerging range of novel imaging systems that would not exist without the AI (and did not exist before the deep learning era). The AI may, for instance, assist in simplifying acquisition and interpretation so non-experts can use an imaging device (such as a low-cost ultrasound device), or extract information from an image that is “beyond the human eye”, or fuse imaging and non-imaging data as a digital phenotype for clinical decision-making.

| Jens Rittscher
| Alison Noble
| Konstantinos Kamnitsas
| Abhirup Banerjee
| Vicente Grau
| Daniel Bulte