The following is a summary of a presentation conducted by our team for the AMIA 2022 Informatics Summit which took place on March 21-24, 2022. For more information, visit https://s4.goeshow.com/amia/summit/2022/schedule_at_a_glance.cfm?session_key=0A46ED0B-9129-113F-C3E4-B3A7CA14E915&session_date=Wednesday,%20Mar%2023,%202022.
Fusion of Chest Radiographs and Electronic Medical Records using Deep Learning to Predict Intubation among Hospitalized Patients with COVID-19
PARENT SESSION
S20: Oral Presentation – Networks, Machine Learning and Other Novel Methods for Clinical Prediction
Description
Both chest radiographs (CXRs) and electronic medical records (EMR), typically obtained early in patients admitted with COVID-19, are key to decide whether patients need intubation. This study evaluates the use of a machine learning model to predict the need for intubation using a combination of CXR and EMR data. The model was trained and validated on historical data. Such a model can prioritize patients with high risk of intubation and result in improving care outcome.