Graphene sensors map vulnerable brain tissue during stroke

Graphene-based sensors could provide a more detailed picture of how brain tissue responds during an ischemic stroke, potentially helping clinicians distinguish tissue that might still be saved from areas where damage has become severe. In experiments in mice, researchers were able to monitor electrical signals that conventional recording technologies struggle to capture.

The international study, published in Brain, involved researchers from the University of Manchester, the Institute of Microelectronics of Barcelona (IMB-CNM, CSIC), the Catalan Institute of Nanoscience and Nanotechnology (ICN2) and Multi Channel Systems in Germany. The technology remains experimental and was tested in animals, but the findings point towards a possible new approach to real-time monitoring of brain vulnerability during stroke.

Capturing a second wave of damage

An ischemic stroke occurs when a blockage interrupts blood flow to part of the brain. Damage can continue after this initial event through cortical spreading depolarizations: waves of abnormal electrical activity that travel across injured brain tissue and can contribute to expansion of the affected area. Studying these events has been difficult because they involve extremely slow electrical changes that conventional monitoring methods cannot reliably capture.

The researchers placed ultrasensitive graphene sensors directly on the brains of mice. This allowed them to record the electrical events in considerably greater detail. Importantly, characteristics of the signals appeared to provide information about the condition of surrounding tissue. The researchers could differentiate between relatively healthy tissue, vulnerable tissue that might potentially recover and areas that had already sustained severe damage. The sensors also revealed how blood flow changed as the electrical waves passed. In healthier regions, blood flow increased, potentially supporting recovery. In vulnerable tissue, however, blood flow sometimes decreased further, potentially adding to the existing injury.

Ketamine altered damaging responses

The researchers subsequently investigated whether these harmful responses could be modified. They administered low doses of ketamine, a drug already used in clinical practice for other indications. In the animal experiments, ketamine shortened the duration of damaging electrical events, improved blood-flow responses and was associated with a reduction in the overall size of brain injuries.

The findings suggest that detailed monitoring could potentially do more than show that harmful electrical activity is occurring. According to the researchers, the characteristics of these signals could provide a real-time indication of which areas of the brain are particularly vulnerable and how tissue is responding to an intervention.

From mice to clinical monitoring

The possible clinical significance lies in identifying brain tissue at risk while there is still an opportunity to intervene. A sufficiently detailed real-time map could eventually help clinicians determine which patients or brain regions might benefit most from additional treatment.

However, substantial steps remain before such an application becomes feasible. The current findings come from mice, with sensors positioned directly on the brain. Whether comparable monitoring can be performed safely and effectively in people with stroke still needs to be established. The ketamine findings should also be interpreted within that experimental context. The study does not demonstrate that administering ketamine to stroke patients will reduce brain damage.

For now, the research primarily shows that graphene sensors can reveal physiological information that has previously been difficult to observe. If subsequent studies translate that capability to humans, monitoring the electrical aftermath of a stroke could provide clinicians with another way to assess not only where damage has occurred, but also where brain tissue may still be salvageable.

Stroke risk prediction

Earlier this year, researchers at Mass General Brigham and the Broad Institute developed ECG2Stroke, an AI model that can predict a patient's stroke risk up to ten years in advance using a standard ECG combined with age and gender. The deep learning model was developed using data from more than 200,000 patients and identifies subtle electrical patterns associated with future stroke risk.

According to the researchers, ECG2Stroke performs comparably to existing clinical risk scores but could be easier to integrate into routine care because it requires no additional testing or complex calculations. The model was particularly effective at predicting cardioembolic strokes, in which blood clots originating in the heart travel to the brain. Abnormalities in the atria played an important role in its predictions. The researchers stress that prospective, real-world validation is still required. If confirmed, ECG2Stroke could help identify high-risk patients earlier and support more targeted preventive measures.