MACHINE LEARNING

  • TRIAL FILES

    We use large language models to summarize practice-changing randomized clinical trials identified on MEDLINE using PaperScrape (a software tool built by our lab).

    Sign up here

  • ER VOLUME

    Prediction Model

    Developing, implementing, and prospectively evaluating a model that accurately forecasts emergency department volumes.

    Am J Emerg Med, 2020

  • HYPOGLYCEMIA

    Prediction Model

    Predicting the risk of hypoglycaemia using machine-learning techniques in hospitalized patients.

    Diabetes Obes Metab, 2021

    Intern Emerg Med, 2022

  • DIALYSIS

    Prediction Model

    Developing a model to predict volume of acute dialysis procedures to optimize nursing resources. Currently in progress.

  • IMAGING

    Prediction Models

    Federated learning was used to read chest x-rays and incorporate clinical data to predict death among patients with COVID-19.

    Nature Medicine, 2021

  • REWARD AI

    Prediction Model

    Creating a decision support tool to identify hospitalized patients at high risk of bleeding sequelae (“high risk”) and evaluating the impact of this tool on rates of critically low hemoglobin and blood transfusion.

    medRxiv, 2025