Mount Sinai Study Says Machine Learning Can Improve Blood Sugar Management
Mount Sinai
|
Contributed by: Kate Gamble
Summary
Researchers at the Icahn School of Medicine at Mount Sinai have created a machine learning tool named GLUCOSE to improve blood sugar management for ICU patients recovering from heart surgery. Published in NPJ Digital Medicine, the study discusses the difficulty in maintaining stable blood sugar levels, as patients are vulnerable to both hyperglycemia and hypoglycemia. GLUCOSE employs reinforcement learning to provide insulin dosing recommendations based on real-time data, enhancing patient safety and supporting clinician decision-making. Initial tests indicate that GLUCOSE can perform on par with experienced clinicians, but the developers underscore its role as an aid rather than a replacement for clinical judgment.
