Revolutionizing Postoperative Ileus Monitoring: Exploring GRU-D's Real-Time Capabilities and Cross-Hospital Transferability

Poster Number: P176
Presentation Time: 05:00 PM - 06:30 PM

Abstract Keywords: Clinical Decision Support, Deep Learning, Critical Care
Primary Track: Applications
Programmatic Theme: Clinical Research Informatics

Postoperative ileus (POI) poses a common challenge following colorectal surgery, contributing to heightened morbidity, increased costs, and prolonged hospital stays. While studies have discussed the prediction of POI under traditional statistical frameworks, there exists a notable gap concerning the performance of deep learning-based approaches in POI prediction. Here we explored the performance and transferability of GRU-D based deep learning architecture for real-time risk modeling of postoperative ileus. Our study indicates strong transferability of the deep learning model across hospital sites and electronic health record systems with non-overlapping surgery date frames. Along with the ability to automate missing parameterization and manage irregular sampling data, the proposed architecture inches closer to real-world clinical practice.

Speaker(s):
Xiaoyang Ruan, PhD
The University of Texas Health Science Center at Houston

Author(s):
Sunyang Fu, PhD, MHI - UTHealth; Kellie Mathis, M.D. - Mayo Clinic; Cornelius Thiels, M.B.A - Mayo Clinic; Patrick Wilson, PhD - Mayo Clinic; Curtis Storlie, PhD - Mayo Clinic; Hongfang Liu, PhD - University of Texas Health Science Center at Houston; Xiaoyang Ruan, PhD - The University of Texas Health Science Center at Houston;