Outcome Prediction in Postanoxic Coma With Deep Learning*
- 1 October 2019
- journal article
- research article
- Published by Ovid Technologies (Wolters Kluwer Health) in Critical Care Medicine
- Vol. 47 (10), 1424-1432
- https://doi.org/10.1097/ccm.0000000000003854
Abstract
Visual assessment of the electroencephalogram by experienced clinical neurophysiologists allows reliable outcome prediction of approximately half of all comatose patients after cardiac arrest. Deep neural networks hold promise to achieve similar or even better performance, being more objective and consistent. Prospective cohort study. Medical ICU of five teaching hospitals in the Netherlands. Eight-hundred ninety-five consecutive comatose patients after cardiac arrest. None. Continuous electroencephalogram was recorded during the first 3 days after cardiac arrest. Functional outcome at 6 months was classified as good (Cerebral Performance Category 1–2) or poor (Cerebral Performance Category 3–5). We trained a convolutional neural network, with a VGG architecture (introduced by the Oxford Visual Geometry Group), to predict neurologic outcome at 12 and 24 hours after cardiac arrest using electroencephalogram epochs and outcome labels as inputs. Output of the network was the probability of good outcome. Data from two hospitals were used for training and internal validation (n = 661). Eighty percent of these data was used for training and cross-validation, the remaining 20% for independent internal validation. Data from the other three hospitals were used for external validation (n = 234). Prediction of poor outcome was most accurate at 12 hours, with a sensitivity in the external validation set of 58% (95% CI, 51–65%) at false positive rate of 0% (CI, 0–7%). Good outcome could be predicted at 12 hours with a sensitivity of 48% (CI, 45–51%) at a false positive rate of 5% (CI, 0–15%) in the external validation set. Deep learning of electroencephalogram signals outperforms any previously reported outcome predictor of coma after cardiac arrest, including visual electroencephalogram assessment by trained electroencephalogram experts. Our approach offers the potential for objective and real time, bedside insight in the neurologic prognosis of comatose patients after cardiac arrest.This publication has 34 references indexed in Scilit:
- A review of unsupervised feature learning and deep learning for time-series modelingPattern Recognition Letters, 2014
- Automated Analysis of Background EEG and Reactivity During Therapeutic Hypothermia in Comatose Patients After Cardiac ArrestClinical Eeg and Neuroscience, 2014
- Predictors of poor neurological outcome in adult comatose survivors of cardiac arrest: A systematic review and meta-analysis. Part 2: Patients treated with therapeutic hypothermiaResuscitation, 2013
- Representation Learning: A Review and New PerspectivesIEEE Transactions on Pattern Analysis and Machine Intelligence, 2013
- A Cerebral Recovery Index (CRI) for early prognosis in patients after cardiac arrestCritical Care, 2013
- Quantitative EEG for the detection of brain ischemiaCritical Care, 2012
- Modeling electroencephalography waveforms with semi-supervised deep belief nets: fast classification and anomaly measurementJournal of Neural Engineering, 2011
- Interobserver Agreement in the Interpretation of EEG Patterns in Critically Ill AdultsJournal of Clinical Neurophysiology, 2008
- Practice Parameter: Prediction of outcome in comatose survivors after cardiopulmonary resuscitation (an evidence-based review)Neurology, 2006
- Über das Elektrenkephalogramm des MenschenArchiv Fur Psychiatrie Und Nervenkrankheiten, 1929