The Microsoft AI team shares research that demonstrates how AI can sequentially investigate and solve medicine’s most complex diagnostic challenges—cases that expert physicians struggle to answer.

Benchmarked against real-world case records published each week in the New England Journal of Medicine, we show that the Microsoft AI Diagnostic Orchestrator (MAI-DxO) correctly diagnoses up to 85% of NEJM case proceedings, a rate more than four times higher than a group of experienced physicians. MAI-DxO also gets to the correct diagnosis more cost-effectively than physicians.

  • technocrit@lemmy.dbzer0.com
    link
    fedilink
    English
    arrow-up
    1
    arrow-down
    2
    ·
    8 hours ago

    The doctor who review the case, maybe ?

    Yeah that’s why these gains in “efficiency” are completely imaginary.

    • boonhet@sopuli.xyz
      link
      fedilink
      English
      arrow-up
      1
      ·
      1 hour ago

      The AI only needs to alert the doctor that something is off and should be tested for. It does not replace doctors, but augments them. It’s actually a great use for AI, it’s just not what we think of as AI in a post-LLM world. The medically useful AI is pattern recognition. LLMs may also help doctors if they need a starting point into researching something weird and obscure, but ChatGPT isn’t being used for diagnosing patients, nor is anything any AI says the “final verdict”. It’s just a tool to improve early detection of disorders, or it might point someone towards an useful article or book.

    • douglasg14b@lemmy.world
      cake
      link
      fedilink
      English
      arrow-up
      3
      ·
      edit-2
      5 hours ago

      Only if you don’t have the critical thinking to understand how information management is a significant problem and barrier to medical care.

      Being able to research and find material relevant to a patient’s problem is an arduous task that often is too high a barrier for doctors to invest in given their regular workloads.

      Which leads to a reduction in effective care.

      By providing a more efficient and effective way to dig up information that saves a ton of time and improves care.

      It’s still up to the doctor to evaluate that information, but now they’re not slogging away trying to find it.