Publication
Imitation learning for clinical decision support in pediatric ECMO
Fateme Golivand Darvishvand; Michael A. Skinner; Saurabh Mathur; Ameet Soni; Phillip Reeder; Kristian Kersting; Lakshmi Raman; Sriraam Natarajan
In: Computing Research Repository eprint Journal (CoRR), Vol. abs/2605.16175, Pages 1-11, a, 2026.
Abstract
Pediatric critical care is a dynamic, high-stakes process in-
volving constant monitoring and adjustments in life-saving treatments.
Modeling these interventions is crucial for effective decision support. To
address the challenges of high complexity and data scarcity in pedi-
atric Extracorporeal Membrane Oxygenation (ECMO), we frame clin-
ical decision-making as learning to act from trajectories, i.e., imitation
learning that learns action models from observational data, with a key
feature that actions are not directly observed. We consider TabPFN,
a recent transformer-based approach for tabular data, and traditional
baselines including XGBoost and Multi-Layer Perceptrons (MLPs) on
real-world pediatric ECMO data to learn the action models. We find
that the TabPFN-based approach consistently outperforms these classi-
cal baselines, supporting its use as a strong clinician-behavior baseline
for pediatric ECMO decision support.