“AI Doctor” Can Predict Medical Outcomes

This post is also available in: עברית (Hebrew)

Artificial intelligence has proven to be useful in reading medical imaging and can even pass doctors’ licensing exams.

This new AI tool takes the challenge to the next level- it can read doctors’ notes and accurately predict patients’ risk of death, hospital readmission, and other important outcomes.

This software was designed by a team at NYU Grossman School of Medicine and is currently in use at the university’s affiliated hospitals throughout New York. A study on its predictive value was published in the journal Nature.

Lead author Eric Oermann, an NYU neurosurgeon and computer scientist, told AFP: “One thing that’s common in medicine everywhere, is physicians write notes about what they’ve seen in clinic, what they’ve discussed with patients… So our basic insight was, can we start with medical notes as our source of data, and then build predictive models on top of it?”

According to Techxplore, a main challenge for the software was interpreting the natural language physicians write in, which varies greatly between individuals.

The language model, called NYUTron, was trained on millions of clinical notes from the health records of 387,000 people who received care within NYU Langone hospitals between January 2011 and May 2020.

NYUTron identified 95% of people who died in a hospital before they were discharged, and 80% of patients who would be readmitted within 30 days. It outperformed most doctors with its predictions, as well as non-AI-based tools.

Nevertheless, Oermann claims AI will never fully substitute a physician-patient relationship. “They will help provide more information for physicians seamlessly at the point-of-care so they can make more informed decisions,” he said.

We may not come to the clinic and be greeted by a stethoscope-wearing robot doctor, but it seems that the integration of AI and human insight is the promising future of medicine.