'AI Doctor' Demonstrates Superior Predictive Ability

Illustration photo: dailymail.co.uk

A new AI tool has demonstrated the ability to read doctors' notes and accurately predict a patient's risk of death, readmission, and other important outcomes in their care.

According to a research report recently published in the journal Nature, the new AI software called NYUTron was developed by a group of scientists at NYU Grossman School of Medicine and is being applied at hospitals affiliated with this university across New York (USA).

The study's lead author, Eric Oermann, a computer scientist and neurosurgeon at NYU, said that while non-AI predictive models have been around in medicine for a long time, they have rarely been used in practice because the data requires reorganization and cumbersome formatting. However, he said, "Doctors often take notes about what they see in the clinic, what they discuss with patients. So our approach was, can we start by using these medical notes as our data source and then build predictive models on top of that?"

The NYUTron large language model was built on millions of clinical notes from the medical records of 387,000 people who were cared for at New York hospitals between January 2011 and May 2020. This data includes everything doctors wrote, such as notes about a patient's progress, X-ray results, and discharge papers. All combined into a corpus of 4.1 billion words.

One of the main challenges for the software is interpreting the natural language that doctors write. Doctors' phrasing varies greatly, including their habitual abbreviations.

By looking back at the records of what happened, the researchers were able to calculate how often NYUTron predicted correctly. Overall, NYUTron correctly identified 95% of patients who died in the hospital before being discharged and 80% of patients who would be readmitted within 30 days. NYUTron also correctly estimated 79% of patients’ actual length of stay, 87% of cases where patients were denied coverage by their insurance, and more. The software outperformed most doctors and non-AI computer models used today.

Mr. Oermann believes that AI will never replace the doctor-patient relationship. However, he asserts that AI can help "provide more information to doctors seamlessly at the point of care, so that professionals can make more informed decisions."

Source link: https://baobinhphuoc.com.vn/news/28/145141/bac-si-ai-chung-minh-kha-nang-du-bao-sieu-viet