Pruning the Path to Optimal Care: Identifying Systematically Suboptimal Medical Decision-Making with Inverse Reinforcement Learning

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Bovenzi, Inko, Carmel, Adi, Hu, Michael, Hurwitz, Rebecca M., McBride, Fiona, Benac, Leo, Ayala, José Roberto Tello, Doshi-Velez, Finale
Format: Preprint
Published: 2024
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866929583525199872
author Bovenzi, Inko
Carmel, Adi
Hu, Michael
Hurwitz, Rebecca M.
McBride, Fiona
Benac, Leo
Ayala, José Roberto Tello
Doshi-Velez, Finale
author_facet Bovenzi, Inko
Carmel, Adi
Hu, Michael
Hurwitz, Rebecca M.
McBride, Fiona
Benac, Leo
Ayala, José Roberto Tello
Doshi-Velez, Finale
contents In aims to uncover insights into medical decision-making embedded within observational data from clinical settings, we present a novel application of Inverse Reinforcement Learning (IRL) that identifies suboptimal clinician actions based on the actions of their peers. This approach centers two stages of IRL with an intermediate step to prune trajectories displaying behavior that deviates significantly from the consensus. This enables us to effectively identify clinical priorities and values from ICU data containing both optimal and suboptimal clinician decisions. We observe that the benefits of removing suboptimal actions vary by disease and differentially impact certain demographic groups.
format Preprint
id arxiv_https___arxiv_org_abs_2411_05237
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Pruning the Path to Optimal Care: Identifying Systematically Suboptimal Medical Decision-Making with Inverse Reinforcement Learning
Bovenzi, Inko
Carmel, Adi
Hu, Michael
Hurwitz, Rebecca M.
McBride, Fiona
Benac, Leo
Ayala, José Roberto Tello
Doshi-Velez, Finale
Machine Learning
Quantitative Methods
Applications
Computation
In aims to uncover insights into medical decision-making embedded within observational data from clinical settings, we present a novel application of Inverse Reinforcement Learning (IRL) that identifies suboptimal clinician actions based on the actions of their peers. This approach centers two stages of IRL with an intermediate step to prune trajectories displaying behavior that deviates significantly from the consensus. This enables us to effectively identify clinical priorities and values from ICU data containing both optimal and suboptimal clinician decisions. We observe that the benefits of removing suboptimal actions vary by disease and differentially impact certain demographic groups.
title Pruning the Path to Optimal Care: Identifying Systematically Suboptimal Medical Decision-Making with Inverse Reinforcement Learning
topic Machine Learning
Quantitative Methods
Applications
Computation
url https://arxiv.org/abs/2411.05237