AI in Healthcare and Medicine
Level 5 · AI you already use
How AI helps doctors and nurses read scans and write notes. Why a scan tool must be cleared before it is sold, and who makes the final call.
Course contentsLesson 13 of 13
Level 1 · Chatbots and voice
Level 2 · What assistants can do
Level 3 · Assistants at work
Level 4 · Search, research and learning
Level 5 · AI you already use
After this lesson you can
- Name the most common job AI does in healthcare
- Explain why a scan reader has to be cleared before it is sold
- Say who makes the final call, and why
9 mincovers 3 glossary entries
Worth reading first: Predictive Text on Phones
How is AI used in healthcare?
AI in healthcare means using AI to help doctors, nurses and hospitals. The most common job is reading medical scans, such as X-rays, and marking spots worth a closer look. The US health agency for drugs and devices lists more than 1,000 AI tools it has cleared. About three in four of them work with scans.
Reading a scan is a computer vision job. A model looks at the picture and marks areas that look like a known problem, such as a small spot on a lung. Sorting each picture into groups, such as 'needs a look' and 'looks normal', is image classification. The doctor then looks at the marked areas first.
Other tools help with paperwork. Some listen to a visit and write up the notes, so the doctor can look at the patient instead of a screen. Others watch a patient's numbers in hospital and warn the nurses when someone may be getting worse.
In the US and the EU, a tool that helps find a disease counts as a medical device. A health agency must clear it before it is sold, and the maker has to show it works. A chatbot that gives general health tips has not been through that. Do not use it as if it had.
A doctor still makes the final call. The AI gives a first pass and marks what to look at. The doctor checks it, adds what they know about the patient, and decides. AI can miss something a doctor would see, and it can mark something that turns out to be nothing.
If you ask a general AI assistant about a health worry, treat the answer as a starting point. It can explain a word on your test results, or help you write questions for your doctor. It cannot examine you, and it may be wrong. Take what it says to a person who can.
Worked example
Lucy reads the morning's X-rays
Lucy, a doctor, has 60 chest X-rays to read before noon. The AI tool marks 7 of them as worth a closer look.
She reads those 7 first and finds 2 real problems. Then she reads the other 53, as she always does, and finds nothing the tool missed.
The words for this lesson3
Check yourself
Who makes the final call when an AI tool marks a spot on a scan?
Finished this lesson?
Marking it complete also marks its 3 glossary entries learned, which updates your roadmap and your saved page. 0 are already marked.