Hemodynamic Instability and Cardiovascular Events after Traumatic Brain Injury Predict Outcome after Artifact Removal with Deep Belief Network Analysis

Hakseung Kim, Seung Bo Lee, Yunsik Son, Marek Czosnyka, Dong Ju Kim

Research output: Contribution to journalArticle

2 Citations (Scopus)

Abstract

Background: Hemodynamic instability and cardiovascular events heavily affect the prognosis of traumatic brain injury. Physiological signals are monitored to detect these events. However, the signals are often riddled with faulty readings, which jeopardize the reliability of the clinical parameters obtained from the signals. A machine-learning model for the elimination of artifactual events shows promising results for improving signal quality. However, the actual impact of the improvements on the performance of the clinical parameters after the elimination of the artifacts is not well studied. Materials and Methods: The arterial blood pressure of 99 subjects with traumatic brain injury was continuously measured for 5 consecutive days, beginning on the day of admission. The machine-learning deep belief network was constructed to automatically identify and remove false incidences of hypotension, hypertension, bradycardia, tachycardia, and alterations in cerebral perfusion pressure (CPP). Results: The prevalences of hypotension and tachycardia were significantly reduced by 47.5% and 13.1%, respectively, after suppressing false incidents (P=0.01). Hypotension was particularly effective at predicting outcome favorability and mortality after artifact elimination (P=0.015 and 0.027, respectively). In addition, increased CPP was also statistically significant in predicting outcomes (P=0.02). Conclusions: The prevalence of false incidents due to signal artifacts can be significantly reduced using machine-learning. Some clinical events, such as hypotension and alterations in CPP, gain particularly high predictive capacity for patient outcomes after artifacts are eliminated from physiological signals.

Original languageEnglish
Pages (from-to)347-353
Number of pages7
JournalJournal of Neurosurgical Anesthesiology
Volume30
Issue number4
DOIs
Publication statusPublished - 2018 Oct 1

Fingerprint

Cerebrovascular Circulation
Hypotension
Artifacts
Hemodynamics
Tachycardia
Bradycardia
Reading
Arterial Pressure
Hypertension
Mortality
Traumatic Brain Injury
Incidence
Machine Learning

Keywords

  • arterial pressure
  • brain injuries
  • computer-assisted
  • machine-learning
  • signal processing
  • traumatic

ASJC Scopus subject areas

  • Surgery
  • Clinical Neurology
  • Anesthesiology and Pain Medicine

Cite this

Hemodynamic Instability and Cardiovascular Events after Traumatic Brain Injury Predict Outcome after Artifact Removal with Deep Belief Network Analysis. / Kim, Hakseung; Lee, Seung Bo; Son, Yunsik; Czosnyka, Marek; Kim, Dong Ju.

In: Journal of Neurosurgical Anesthesiology, Vol. 30, No. 4, 01.10.2018, p. 347-353.

Research output: Contribution to journalArticle

@article{950c086f68394df3b38bfcbf41ce9ab6,
title = "Hemodynamic Instability and Cardiovascular Events after Traumatic Brain Injury Predict Outcome after Artifact Removal with Deep Belief Network Analysis",
abstract = "Background: Hemodynamic instability and cardiovascular events heavily affect the prognosis of traumatic brain injury. Physiological signals are monitored to detect these events. However, the signals are often riddled with faulty readings, which jeopardize the reliability of the clinical parameters obtained from the signals. A machine-learning model for the elimination of artifactual events shows promising results for improving signal quality. However, the actual impact of the improvements on the performance of the clinical parameters after the elimination of the artifacts is not well studied. Materials and Methods: The arterial blood pressure of 99 subjects with traumatic brain injury was continuously measured for 5 consecutive days, beginning on the day of admission. The machine-learning deep belief network was constructed to automatically identify and remove false incidences of hypotension, hypertension, bradycardia, tachycardia, and alterations in cerebral perfusion pressure (CPP). Results: The prevalences of hypotension and tachycardia were significantly reduced by 47.5{\%} and 13.1{\%}, respectively, after suppressing false incidents (P=0.01). Hypotension was particularly effective at predicting outcome favorability and mortality after artifact elimination (P=0.015 and 0.027, respectively). In addition, increased CPP was also statistically significant in predicting outcomes (P=0.02). Conclusions: The prevalence of false incidents due to signal artifacts can be significantly reduced using machine-learning. Some clinical events, such as hypotension and alterations in CPP, gain particularly high predictive capacity for patient outcomes after artifacts are eliminated from physiological signals.",
keywords = "arterial pressure, brain injuries, computer-assisted, machine-learning, signal processing, traumatic",
author = "Hakseung Kim and Lee, {Seung Bo} and Yunsik Son and Marek Czosnyka and Kim, {Dong Ju}",
year = "2018",
month = "10",
day = "1",
doi = "10.1097/ANA.0000000000000462",
language = "English",
volume = "30",
pages = "347--353",
journal = "Journal of Neurosurgical Anesthesiology",
issn = "0898-4921",
publisher = "Lippincott Williams and Wilkins",
number = "4",

}

TY - JOUR

T1 - Hemodynamic Instability and Cardiovascular Events after Traumatic Brain Injury Predict Outcome after Artifact Removal with Deep Belief Network Analysis

AU - Kim, Hakseung

AU - Lee, Seung Bo

AU - Son, Yunsik

AU - Czosnyka, Marek

AU - Kim, Dong Ju

PY - 2018/10/1

Y1 - 2018/10/1

N2 - Background: Hemodynamic instability and cardiovascular events heavily affect the prognosis of traumatic brain injury. Physiological signals are monitored to detect these events. However, the signals are often riddled with faulty readings, which jeopardize the reliability of the clinical parameters obtained from the signals. A machine-learning model for the elimination of artifactual events shows promising results for improving signal quality. However, the actual impact of the improvements on the performance of the clinical parameters after the elimination of the artifacts is not well studied. Materials and Methods: The arterial blood pressure of 99 subjects with traumatic brain injury was continuously measured for 5 consecutive days, beginning on the day of admission. The machine-learning deep belief network was constructed to automatically identify and remove false incidences of hypotension, hypertension, bradycardia, tachycardia, and alterations in cerebral perfusion pressure (CPP). Results: The prevalences of hypotension and tachycardia were significantly reduced by 47.5% and 13.1%, respectively, after suppressing false incidents (P=0.01). Hypotension was particularly effective at predicting outcome favorability and mortality after artifact elimination (P=0.015 and 0.027, respectively). In addition, increased CPP was also statistically significant in predicting outcomes (P=0.02). Conclusions: The prevalence of false incidents due to signal artifacts can be significantly reduced using machine-learning. Some clinical events, such as hypotension and alterations in CPP, gain particularly high predictive capacity for patient outcomes after artifacts are eliminated from physiological signals.

AB - Background: Hemodynamic instability and cardiovascular events heavily affect the prognosis of traumatic brain injury. Physiological signals are monitored to detect these events. However, the signals are often riddled with faulty readings, which jeopardize the reliability of the clinical parameters obtained from the signals. A machine-learning model for the elimination of artifactual events shows promising results for improving signal quality. However, the actual impact of the improvements on the performance of the clinical parameters after the elimination of the artifacts is not well studied. Materials and Methods: The arterial blood pressure of 99 subjects with traumatic brain injury was continuously measured for 5 consecutive days, beginning on the day of admission. The machine-learning deep belief network was constructed to automatically identify and remove false incidences of hypotension, hypertension, bradycardia, tachycardia, and alterations in cerebral perfusion pressure (CPP). Results: The prevalences of hypotension and tachycardia were significantly reduced by 47.5% and 13.1%, respectively, after suppressing false incidents (P=0.01). Hypotension was particularly effective at predicting outcome favorability and mortality after artifact elimination (P=0.015 and 0.027, respectively). In addition, increased CPP was also statistically significant in predicting outcomes (P=0.02). Conclusions: The prevalence of false incidents due to signal artifacts can be significantly reduced using machine-learning. Some clinical events, such as hypotension and alterations in CPP, gain particularly high predictive capacity for patient outcomes after artifacts are eliminated from physiological signals.

KW - arterial pressure

KW - brain injuries

KW - computer-assisted

KW - machine-learning

KW - signal processing

KW - traumatic

UR - http://www.scopus.com/inward/record.url?scp=85048769981&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85048769981&partnerID=8YFLogxK

U2 - 10.1097/ANA.0000000000000462

DO - 10.1097/ANA.0000000000000462

M3 - Article

C2 - 28991060

AN - SCOPUS:85048769981

VL - 30

SP - 347

EP - 353

JO - Journal of Neurosurgical Anesthesiology

JF - Journal of Neurosurgical Anesthesiology

SN - 0898-4921

IS - 4

ER -