Artificial Intelligence in Medicine
Also called: medical AI, clinical AI
Software that performs tasks in healthcare which normally require human judgment, most often pattern recognition in images, text, or signals.
Artificial intelligence in medicine describes computational systems that carry out tasks normally requiring clinical judgment: detecting a finding on an image, estimating a risk, drafting a note, or answering a clinical question. Nearly all of it in current practice is machine learning, meaning the behavior was learned from data rather than programmed as explicit rules.
The practical categories are narrower than the term suggests. Most authorized medical AI does one thing on one data type for one clinical question. A device that triages head CT for hemorrhage cannot triage anything else, and calling it intelligent obscures how specific it actually is.
Where This Gets Misread
The word invites a comparison to human cognition that the technology does not support. These are statistical pattern recognizers with narrow competence and no model of the patient. The useful question is never whether a system is intelligent, it is what task it was validated on and in whom.