Medical

EEG scan detects hidden consciousness in unresponsive brain injury patients

EEG scan detects hidden consci...
Using machine learning, scientists could identify patterns of hidden consciousness in EEG data from unresponsive patients, and predict their likelihood of recovery
Using machine learning, scientists could identify patterns of hidden consciousness in EEG data from unresponsive patients, and predict their likelihood of recovery
View 1 Image
Using machine learning, scientists could identify patterns of hidden consciousness in EEG data from unresponsive patients, and predict their likelihood of recovery
1/1
Using machine learning, scientists could identify patterns of hidden consciousness in EEG data from unresponsive patients, and predict their likelihood of recovery

A landmark study from neurologists at Columbia University hasdescribed a new method to detect signs ofconsciousness in unresponsive brain-injured patients using a simple EEG scan. The researchalso suggests the EEG data can predict which patients have a highchance of regaining consciousness and recovering.

About a decade ago advances in medical imaging allowed scientiststo better track deep signals of consciousness in brain-injuredcomatose patients. MRI data revealed what was dubbed "hiddenconsciousness", sometimes also referred to as covert consciousness.

This undeniable medical breakthrough broadened the definition ofconsciousness for academic purposes, but several challenges remainedbefore this knowledge could be effectively translated into something clinicallyfunctional. For one, MRI tests are not easy to perform in intensivecare situations, and even then, it was unclear exactly whatthese signs of hidden consciousness actually meant in terms ofpatient recovery.

"One of the most challenging problems in the ICU involvespredicting recovery, and not just survival, for patients who areunconscious after a brain injury," says Jan Claassen, lead authoron the new research. "Since the first studies describing hiddenconsciousness, we've been looking for a practical way to do this inthe early days after brain injury, when treatment decisions thataffect outcomes are often made."

The new study followed 104 patients in intensive care aftersuffering major brain injury. All the subjects were completelyunresponsive, but not paralyzed. While EEG scans were conducted, thesubjects were asked to open and close their hands. A machine learningalgorithm was then tasked with analyzing the EEG scans to ascertainwhether the physical commands triggered any tiny patterns in theneurological data.

The results revealed unique EEG patterns in response to thecommands in 15 percent of the unresponsive subjects within the firstfour days of the brain injury. Around half of patients identifiedwith these signs of hidden consciousness improved to the point theycould follow verbal commands by the time they were discharged fromhospital. Only 26 percent of subjects without these initial signsimproved to the same point.

A year later 44 percent of the group displaying the initial EEGactivity were independently functional for up to eight hours everyday. This compares to only 14 percent reaching the same point a yearlater without initially displaying these EEG activity signs.

Although these results are incredibly promising, much more work isstill needed before they are translated into clinical applications.The study was not large enough to separate out different kinds ofbrain injuries. Early indications suggest the EEG test is mosteffective in cases of brain haemorrhage and trauma, as opposed toinjury suffered from oxygen deprivation, but this will need to beclarified through larger trials.

Claassen also says there are some technical hurdles that needto be overcome before this test is broadly implemented. Computers areneeded to process the EEG data, and because patients tend to drift inand out of consciousness after a brain injury, the tests need to beconducted several times a day in the first few days following aninjury for the most accurate results.

"It's important to begin monitoring with EEG as early aspossible and to assess at several time points as recovery after asevere brain injury is a complex process," says Claassen.

Still, this new discovery points to potentially groundbreakingchanges in how clinicians approach unresponsive patients in the firstfew days following brain injury. If the technique can be effectively fine-tuned, it could be easily rolled out into intensive care unitsaround the world and assist doctors to better target those patients mostlikely to recover consciousness following a brain injury.

"Though our study was small, it suggests that EEG – a toolthat's readily available at the patient's bedside in the ICU inalmost any hospital across the globe – has the potential tocompletely change how we manage patients with acute brain injury,"says Claassen.

The research was published in The New England Journal of Medicine.

Source: Columbia University

2 comments
Edward Vix
More evidence against the concept of "brain death" which has been used for several decades now by transplant surgeons to subjectively select young healthy organ donors, especially of heart, liver and lung, and then remove these organs while the victims are un-anaensthetized, because that works best for the organs. Common sense says that if the heart is beating on its own and begins to race when the first incision is made, the victim is responding in pain and the action of removing the heart, for example, is what really causes death. This is kind of like what the Aztecs used to do to make sure the sun came up the next day. Just saying.
W Alan Jones
Autonomic response explains racing of heart. Your use of the word 'Victim" shows where you're coming from.