Nieke Roos
6 June 2019

Researchers of the University of Twente and the Medisch Spectrum Twente hospital developed a learning network that’s capable of interpreting the EEG patterns of coma patients, providing better insight into their prospects. Artificial Intelligence (AI) can give a reliable outcome prediction, and thus forms a valuable extra source of information for decision-making. The researchers present their approach in the Critical Care Medicine journal.

In the Netherlands, about one-third of the people that had a cardiac arrest followed by resuscitation will have to be treated in the intensive care unit. These patients, about 7,000 each year, are in a coma. More than half of them will not regain consciousness. The family will want to know what the prospects are and, if their relative regains consciousness, what will be the quality of life. The question “Does further treatment make sense?” can only be answered after careful analysis of the situation.


The brain’s electrical signals – the EEG patterns measured via electrodes on the head – provide a lot of information. EEG analysis using AI gives a very accurate outcome prediction, as the Dutch researchers now show in their paper. Twelve hours after resuscitation, the learning network is capable of predicting a good outcome with 58 percent accuracy and a bad outcome with 48 percent. This is a better performance than the trained eye of a neurologist. Both computer and human, however, still have a category ‘I don’t know’, in situations the EEG data are not specific enough.

The deep learning network has been trained using 600 EEG patterns, without getting any hints on what to look at. After that, it was fed with 300 EEG patterns to see how it performed in giving a prediction. Neurologists have to look at hundreds of EEGs as well, as part of their training. An experienced neurologist will guide them and point out what they have to look at. Still, the EEG patterns are so information rich that the computer outperforms human judgment.

Once trained, the network will be capable of judging the EEG very quickly, well within a second. The researchers expect that this adds valuable information to human judgment. One of the other advantages is flexibility: analyses can be made at any time of the day. Using the new technology at ICUs will have to make clear if the ‘intensivist’ also sees it as a valuable tool. One of the next research steps is having a closer look at the network’s learning strategy, making it more transparent than a black box approach. For this, the neurophysiologists collaborate with computer scientists and mathematicians of the University of Twente.