Predicting neurological outcome after out-of-hospital cardiac arrest with cumulative information; development and internal validation of an artificial neural network algorithm
Abstract Background Prognostication of neurological outcome in patients who remain comatose after cardiac arrest resuscitation is complex. Clinical variables, as well as biomarkers of brain injury, cardiac injury, and systemic inflammation, all yield some prognostic value. We hypothesised that cumul...
Main Authors: | , , , , , , , , , , , , , , , |
---|---|
Format: | Article |
Language: | English |
Published: |
BMC
2021-02-01
|
Series: | Critical Care |
Subjects: | |
Online Access: | https://doi.org/10.1186/s13054-021-03505-9 |
id |
doaj-861198eaad274603985dbad360bf78f7 |
---|---|
record_format |
Article |
collection |
DOAJ |
language |
English |
format |
Article |
sources |
DOAJ |
author |
Peder Andersson Jesper Johnsson Ola Björnsson Tobias Cronberg Christian Hassager Henrik Zetterberg Pascal Stammet Johan Undén Jesper Kjaergaard Hans Friberg Kaj Blennow Gisela Lilja Matt P. Wise Josef Dankiewicz Niklas Nielsen Attila Frigyesi |
spellingShingle |
Peder Andersson Jesper Johnsson Ola Björnsson Tobias Cronberg Christian Hassager Henrik Zetterberg Pascal Stammet Johan Undén Jesper Kjaergaard Hans Friberg Kaj Blennow Gisela Lilja Matt P. Wise Josef Dankiewicz Niklas Nielsen Attila Frigyesi Predicting neurological outcome after out-of-hospital cardiac arrest with cumulative information; development and internal validation of an artificial neural network algorithm Critical Care Machine learning Artificial intelligence Artificial neural networks Neural networks Out-of-hospital cardiac arrest Cardiac arrest |
author_facet |
Peder Andersson Jesper Johnsson Ola Björnsson Tobias Cronberg Christian Hassager Henrik Zetterberg Pascal Stammet Johan Undén Jesper Kjaergaard Hans Friberg Kaj Blennow Gisela Lilja Matt P. Wise Josef Dankiewicz Niklas Nielsen Attila Frigyesi |
author_sort |
Peder Andersson |
title |
Predicting neurological outcome after out-of-hospital cardiac arrest with cumulative information; development and internal validation of an artificial neural network algorithm |
title_short |
Predicting neurological outcome after out-of-hospital cardiac arrest with cumulative information; development and internal validation of an artificial neural network algorithm |
title_full |
Predicting neurological outcome after out-of-hospital cardiac arrest with cumulative information; development and internal validation of an artificial neural network algorithm |
title_fullStr |
Predicting neurological outcome after out-of-hospital cardiac arrest with cumulative information; development and internal validation of an artificial neural network algorithm |
title_full_unstemmed |
Predicting neurological outcome after out-of-hospital cardiac arrest with cumulative information; development and internal validation of an artificial neural network algorithm |
title_sort |
predicting neurological outcome after out-of-hospital cardiac arrest with cumulative information; development and internal validation of an artificial neural network algorithm |
publisher |
BMC |
series |
Critical Care |
issn |
1364-8535 |
publishDate |
2021-02-01 |
description |
Abstract Background Prognostication of neurological outcome in patients who remain comatose after cardiac arrest resuscitation is complex. Clinical variables, as well as biomarkers of brain injury, cardiac injury, and systemic inflammation, all yield some prognostic value. We hypothesised that cumulative information obtained during the first three days of intensive care could produce a reliable model for predicting neurological outcome following out-of-hospital cardiac arrest (OHCA) using artificial neural network (ANN) with and without biomarkers. Methods We performed a post hoc analysis of 932 patients from the Target Temperature Management trial. We focused on comatose patients at 24, 48, and 72 h post-cardiac arrest and excluded patients who were awake or deceased at these time points. 80% of the patients were allocated for model development (training set) and 20% for internal validation (test set). To investigate the prognostic potential of different levels of biomarkers (clinically available and research-grade), patients’ background information, and intensive care observation and treatment, we created three models for each time point: (1) clinical variables, (2) adding clinically accessible biomarkers, e.g., neuron-specific enolase (NSE) and (3) adding research-grade biomarkers, e.g., neurofilament light (NFL). Patient outcome was the dichotomised Cerebral Performance Category (CPC) at six months; a good outcome was defined as CPC 1–2 whilst a poor outcome was defined as CPC 3–5. The area under the receiver operating characteristic curve (AUROC) was calculated for all test sets. Results AUROC remained below 90% when using only clinical variables throughout the first three days in the ICU. Adding clinically accessible biomarkers such as NSE, AUROC increased from 82 to 94% (p < 0.01). The prognostic accuracy remained excellent from day 1 to day 3 with an AUROC at approximately 95% when adding research-grade biomarkers. The models which included NSE after 72 h and NFL on any of the three days had a low risk of false-positive predictions while retaining a low number of false-negative predictions. Conclusions In this exploratory study, ANNs provided good to excellent prognostic accuracy in predicting neurological outcome in comatose patients post OHCA. The models which included NSE after 72 h and NFL on all days showed promising prognostic performance. |
topic |
Machine learning Artificial intelligence Artificial neural networks Neural networks Out-of-hospital cardiac arrest Cardiac arrest |
url |
https://doi.org/10.1186/s13054-021-03505-9 |
work_keys_str_mv |
AT pederandersson predictingneurologicaloutcomeafteroutofhospitalcardiacarrestwithcumulativeinformationdevelopmentandinternalvalidationofanartificialneuralnetworkalgorithm AT jesperjohnsson predictingneurologicaloutcomeafteroutofhospitalcardiacarrestwithcumulativeinformationdevelopmentandinternalvalidationofanartificialneuralnetworkalgorithm AT olabjornsson predictingneurologicaloutcomeafteroutofhospitalcardiacarrestwithcumulativeinformationdevelopmentandinternalvalidationofanartificialneuralnetworkalgorithm AT tobiascronberg predictingneurologicaloutcomeafteroutofhospitalcardiacarrestwithcumulativeinformationdevelopmentandinternalvalidationofanartificialneuralnetworkalgorithm AT christianhassager predictingneurologicaloutcomeafteroutofhospitalcardiacarrestwithcumulativeinformationdevelopmentandinternalvalidationofanartificialneuralnetworkalgorithm AT henrikzetterberg predictingneurologicaloutcomeafteroutofhospitalcardiacarrestwithcumulativeinformationdevelopmentandinternalvalidationofanartificialneuralnetworkalgorithm AT pascalstammet predictingneurologicaloutcomeafteroutofhospitalcardiacarrestwithcumulativeinformationdevelopmentandinternalvalidationofanartificialneuralnetworkalgorithm AT johanunden predictingneurologicaloutcomeafteroutofhospitalcardiacarrestwithcumulativeinformationdevelopmentandinternalvalidationofanartificialneuralnetworkalgorithm AT jesperkjaergaard predictingneurologicaloutcomeafteroutofhospitalcardiacarrestwithcumulativeinformationdevelopmentandinternalvalidationofanartificialneuralnetworkalgorithm AT hansfriberg predictingneurologicaloutcomeafteroutofhospitalcardiacarrestwithcumulativeinformationdevelopmentandinternalvalidationofanartificialneuralnetworkalgorithm AT kajblennow predictingneurologicaloutcomeafteroutofhospitalcardiacarrestwithcumulativeinformationdevelopmentandinternalvalidationofanartificialneuralnetworkalgorithm AT giselalilja predictingneurologicaloutcomeafteroutofhospitalcardiacarrestwithcumulativeinformationdevelopmentandinternalvalidationofanartificialneuralnetworkalgorithm AT mattpwise predictingneurologicaloutcomeafteroutofhospitalcardiacarrestwithcumulativeinformationdevelopmentandinternalvalidationofanartificialneuralnetworkalgorithm AT josefdankiewicz predictingneurologicaloutcomeafteroutofhospitalcardiacarrestwithcumulativeinformationdevelopmentandinternalvalidationofanartificialneuralnetworkalgorithm AT niklasnielsen predictingneurologicaloutcomeafteroutofhospitalcardiacarrestwithcumulativeinformationdevelopmentandinternalvalidationofanartificialneuralnetworkalgorithm AT attilafrigyesi predictingneurologicaloutcomeafteroutofhospitalcardiacarrestwithcumulativeinformationdevelopmentandinternalvalidationofanartificialneuralnetworkalgorithm |
_version_ |
1724225189437243392 |
spelling |
doaj-861198eaad274603985dbad360bf78f72021-03-11T11:48:32ZengBMCCritical Care1364-85352021-02-0125111210.1186/s13054-021-03505-9Predicting neurological outcome after out-of-hospital cardiac arrest with cumulative information; development and internal validation of an artificial neural network algorithmPeder Andersson0Jesper Johnsson1Ola Björnsson2Tobias Cronberg3Christian Hassager4Henrik Zetterberg5Pascal Stammet6Johan Undén7Jesper Kjaergaard8Hans Friberg9Kaj Blennow10Gisela Lilja11Matt P. Wise12Josef Dankiewicz13Niklas Nielsen14Attila Frigyesi15Department of Clinical Sciences Lund, Anaesthesia and Intensive Care, Lund University, Skåne University HospitalDepartment of Clinical Sciences Lund, Anesthesia and Intensive Care, Lund University, Helsingborg HospitalDepartment of Energy Sciences, Faculty of Engineering, Lund UniversityDepartment of Clinical Sciences Lund, Neurology, Lund University, Skåne University HospitalDepartment of Cardiology, Rigshospitalet and Department of Clinical Medicine, University of CopenhagenDepartment of Psychiatry and Neurochemistry, Institute of Neuroscience and Physiology, The Sahlgrenska Academy At the University of GothenburgMedical and Health Directorate, National Fire and Rescue CorpsDepartment of Clinical Sciences Malmö, Anaesthesia and Intensive Care, Lund University, Hallands Hospital HalmstadDepartment of Cardiology, Rigshospitalet and Department of Clinical Medicine, University of CopenhagenDepartment of Clinical Sciences Lund, Anaesthesia and Intensive Care, Lund University, Skåne University HospitalDepartment of Psychiatry and Neurochemistry, Institute of Neuroscience and Physiology, The Sahlgrenska Academy At the University of GothenburgDepartment of Clinical Sciences Lund, Neurology, Lund University, Skåne University HospitalAdult Critical Care, University Hospital of WalesDepartment of Clinical Sciences Lund, Cardiology, Lund University, Skåne University HospitalDepartment of Clinical Sciences Lund, Anesthesia and Intensive Care, Lund University, Helsingborg HospitalDepartment of Clinical Sciences Lund, Anaesthesia and Intensive Care, Lund University, Skåne University HospitalAbstract Background Prognostication of neurological outcome in patients who remain comatose after cardiac arrest resuscitation is complex. Clinical variables, as well as biomarkers of brain injury, cardiac injury, and systemic inflammation, all yield some prognostic value. We hypothesised that cumulative information obtained during the first three days of intensive care could produce a reliable model for predicting neurological outcome following out-of-hospital cardiac arrest (OHCA) using artificial neural network (ANN) with and without biomarkers. Methods We performed a post hoc analysis of 932 patients from the Target Temperature Management trial. We focused on comatose patients at 24, 48, and 72 h post-cardiac arrest and excluded patients who were awake or deceased at these time points. 80% of the patients were allocated for model development (training set) and 20% for internal validation (test set). To investigate the prognostic potential of different levels of biomarkers (clinically available and research-grade), patients’ background information, and intensive care observation and treatment, we created three models for each time point: (1) clinical variables, (2) adding clinically accessible biomarkers, e.g., neuron-specific enolase (NSE) and (3) adding research-grade biomarkers, e.g., neurofilament light (NFL). Patient outcome was the dichotomised Cerebral Performance Category (CPC) at six months; a good outcome was defined as CPC 1–2 whilst a poor outcome was defined as CPC 3–5. The area under the receiver operating characteristic curve (AUROC) was calculated for all test sets. Results AUROC remained below 90% when using only clinical variables throughout the first three days in the ICU. Adding clinically accessible biomarkers such as NSE, AUROC increased from 82 to 94% (p < 0.01). The prognostic accuracy remained excellent from day 1 to day 3 with an AUROC at approximately 95% when adding research-grade biomarkers. The models which included NSE after 72 h and NFL on any of the three days had a low risk of false-positive predictions while retaining a low number of false-negative predictions. Conclusions In this exploratory study, ANNs provided good to excellent prognostic accuracy in predicting neurological outcome in comatose patients post OHCA. The models which included NSE after 72 h and NFL on all days showed promising prognostic performance.https://doi.org/10.1186/s13054-021-03505-9Machine learningArtificial intelligenceArtificial neural networksNeural networksOut-of-hospital cardiac arrestCardiac arrest |