In a recent study, Microsoft researchers revealed that a new AI tool capable of diagnosing patients in a step-by-step manner, similar to how a doctor would, proved to be over twice as accurate as general physicians in handling complex medical cases.
Published on Thursday in Science, the study outlined that this AI model distinguishes itself from typical diagnostic tools by adopting a unique methodology.
According to Microsoft, the AI identifies relevant questions to ask patients while also determining necessary exams and tests, utilizing what they describe as an “orchestrator.” This orchestrator essentially organizes how the AI approaches solving diagnostic puzzles by simulating a panel of clinical experts working together, the report indicates.
In contrast, most existing AI models tend to use a broad approach, analyzing all available patient data—such as symptoms, lab results, and imaging studies—simultaneously to produce a diagnosis.
MICROSOFT PRESIDENT SAYS AI FACES FEWER SAFETY RULES THAN MILK IN CALL FOR GLOBAL REGULATION
While the orchestrator manages the diagnostic progression, the AI starts by asking pertinent questions related to the patient’s medical history, all while pinpointing critical findings during physical examinations.
As the process unfolds, it decides on which tests to order and adjusts based on the results, continuously refining its hypotheses and diagnostic plans.
MICROSOFT UNVEILS CODE OF CONDUCT FOR AI MODELS AS SAFETY CONCERNS MOUNT
This tool outperformed human medical practitioners when faced with various scenarios taken from a large database of emergency room admissions. Researchers pointed out that many older system assessments compare their performance against human doctors using multiple-choice questions.
Dr. Dominic King, the corporate VP of the health unit at Microsoft AI, noted that while AI has excelled at achieving perfect scores in those tests, it doesn’t reflect real-life situations. After all, no patient arrives with neatly summarized symptoms.
The researchers challenged the AI model with over 300 unusual and complicated medical cases, finding that the orchestration method significantly enhanced accuracy in diagnosing patients compared to general physicians relying on textbooks and the internet.
ANTHROPIC SILENT ON WHETHER IT WILL DISTRIBUTE SHARES TO AMERICANS FOLLOWING MUSK SPACEX PLEDGE
Furthermore, the findings indicated that diagnostic processes became less expensive when the orchestrator directed the work-ups compared to when human doctors conducted them.
Harsha Nori, head of AI at Microsoft AI Health, explained, “We’re not just measuring the accuracy of the systems. We’re also looking at how judiciously and thoroughly it goes through a diagnostic work-up.”
The next phase for the Microsoft research team will involve collaborating with Beth Israel Deaconess Medical Center in Boston to assess the tool’s effectiveness in triaging patients within emergency department waiting rooms.
