An Anytime Querying Algorithm for Predicting Cardiac Arrest in Children: Work-in-Progress

Michael A. Skinner, Priscilla Yu, Lakshmi Raman, Sriraam Natarajan

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

Cardiac arrest (CA) is a devastating complication for children in the cardiac intensive care unit (CICU). We developed an “anytime" algorithm to predict CA, using the first few hours of EHR data for initial approximation, and then using information from subsequent time periods to augment the predictive model, improving performance at each iteration. Our initial empirical evaluation on EHR CICU data shows that the model achieves significantly higher performance than learning with all the available data at each iteration when predicting CA inside CICU.

Original languageEnglish (US)
Title of host publicationArtificial Intelligence in Medicine - 20th International Conference on Artificial Intelligence in Medicine, AIME 2022, Proceedings
EditorsMartin Michalowski, Syed Sibte Raza Abidi, Samina Abidi
PublisherSpringer Science and Business Media Deutschland GmbH
Pages353-357
Number of pages5
ISBN (Print)9783031093418
DOIs
StatePublished - 2022
Event20th International Conference on Artificial Intelligence in Medicine, AIME 2022 - Halifax, Canada
Duration: Jun 14 2022Jun 17 2022

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume13263 LNAI
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Conference

Conference20th International Conference on Artificial Intelligence in Medicine, AIME 2022
Country/TerritoryCanada
CityHalifax
Period6/14/226/17/22

Keywords

  • Anytime algorithms
  • Cardiac arrest
  • Gradient boosting

ASJC Scopus subject areas

  • Theoretical Computer Science
  • General Computer Science

Fingerprint

Dive into the research topics of 'An Anytime Querying Algorithm for Predicting Cardiac Arrest in Children: Work-in-Progress'. Together they form a unique fingerprint.

Cite this