AI in HealthCare

Artificial intelligence is now part of healthcare. Hospitals, health systems, and life science companies are building AI tools, and most of them fall into three jobs: helping clinicians spot problems in scans and test results, supporting decisions about patient care, and cutting down on paperwork and other administrative work.

Will AI replace doctors?

No credible source expects AI to replace doctors or nurses any time soon. Healthcare involves judgment, communication, and responsibility that software cannot take over. What is happening instead is that AI is becoming a second set of eyes. Some AI systems have shown strong results at spotting certain findings in medical images, such as suspicious spots on scans, but results vary by task and hospital, and a clinician still reviews and signs off. Think of AI as an assistant that flags things, not a replacement that decides.

“‘Since the start of the COVID pandemic, the number of incoming electronic messages to our providers has increased by 57%. [So we’re piloting] a generative AI tool that turns incoming messages into editable drafts. ”
— Quote SourceFrank Liao, PhD, director of AI, emerging technologies and software engineering at UW Health in Madison, Wisc. (source: WISN-TV)

How AI learns from data

Most medical AI relies on machine learning, which means a computer finds patterns in many examples instead of following hand-written rules. Show a system thousands of labeled scans, and it can learn which patterns tend to go with a certain condition. The quality of the data matters a great deal. If the training data leaves out certain groups of patients, the tool can work worse for them, which is one of the main ethical concerns in medical AI.

Things to keep in mind

  • Patient privacy has to be protected whenever health data is used to train or run an AI tool.

  • AI can be wrong, so a trained professional should always check important results.

  • Tools should be tested on many different kinds of patients before they are trusted.

Want to go deeper? Try our free AI in Medicine course.

Sources: National Institutes of Health (nih.gov) materials on AI in health care, and UW Health reporting quoted above (WISN-TV).

For more, watch this TED talk from Navid Toosy Saidy.

Previous
Previous

Learn About AI Tools: ChatGPT