Outcome Prediction in Postanoxic Coma With Deep Learning*

Abstract

Objectives: 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. Design: Prospective cohort study. Setting: Medical ICU of five teaching hospitals in the Netherlands. Patients: Eight-hundred ninety-five consecutive comatose patients after cardiac arrest. Interventions: None.Measurements and Main Results: 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. Conclusions: 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.

Quantitative and Qualitative EEG as a Prediction Tool for Outcome and Complications in Acute Stroke Patients

“…Machine-learning algorithms could further facilitate outcome prediction using EEG, as has already been shown in the outcome prediction of comatose patients after cardiac arrest. 62…”

Proceedings from the Neurotherapeutics Symposium on Neurological Emergencies: Shaping the Future of Neurocritical Care

“…This state known as "cognitive motor dissociation" was previously demonstrated in chronically unresponsive patients [80], and more recently in acute brain injury [78,81]. Similarly, the use of machine learning and quantitative analysis of longitudinal EEG signals shows promise in improving accuracy of commonly employed neurophysiologic predictors of outcome after cardiac arrest, such as background reactivity, epileptiform activity, and background continuity [79,82].…”

Big data, machine learning and artificial intelligence: a neurologist’s guide

“…A deep-learning artificial neural network analysing EEG data in comatose patients 12 h following cardiac arrest was able to predict 6-month functional outcome: good (48% accuracy and 0% falsepositive rate) versus bad (58% accuracy and 5% falsepositive rate). 17 Work in this area has also begun to go beyond passive recording and classification of inputs with the development of brain-machine interfaces that offer the intriguing potential to one day to communicate with people in a coma. 18…”

“…A deep-learning artificial neural network analysing EEG data in comatose patients 12 h following cardiac arrest was able to predict 6-month functional outcome: good (48% accuracy and 0% false-positive rate) versus bad (58% accuracy and 5% false-positive rate). 17…”