According to the WHO, 15 million people suffer from stroke each year, of which 5 million succumb to fatal injuries, and another 5 million are permanently disabled, making stroke the second highest cause of disability worldwide. The disease is caused by a lack of blood flow to areas of the brain or spinal cord, either by an occlusion of a large artery (ischemic stroke), or as a downstream effect of a bleed into the brain (hemorrhagic stroke). The current treatment for an ischemic stroke is to give medication called tissue plasminogen activator (tPA). A clot buster, tPA can act to dissolve blood clots and restore circulation to the affected area of the brain. While it’s been shown to work highly effectively when used for ischemic stroke, it can cause more bleeding and worsen outcomes in the case of hemorrhagic stroke. Thus, the workflow to diagnose the type of stroke is an important step in its management. However, the process is lengthy and includes evaluation by stroke teams and long wait times for CT scans. It’s been thought that 1.9 million neurons die every minute that a stroke goes untreated, so timing is critical (as neurologists say: “Time is brain!”). Making things more complicated is the fact that tPA treatment is only shown to be effective within a 4.5-hour window, starting from the onset of symptoms.
Scientists in Sweden, from Chalmers University of Technology, Sahlgrenska Academy, and Sahlgrenska University Hospital, have created a new helmet, called Strokefinder, for the rapid identification of hemorrhagic stroke. The idea is that such helmets would be equipped into ambulances and aide in the diagnosis of patients on their way to the hospital to shave precious minutes before treatment can begin. The helmet detects the presence of pooled blood in the brain by sending and receiving biologically-safe microwave pulses (no, there’s no “Popcorn” setting on the helmet). If a bleed is present, the helmet will pick up the signal as a unique scatter pattern.
The system was made to teach itself which microwave scatter patterns point to hemorrhagic stroke via a computer learning algorithm and has shown considerable results, given the small sample size (roughly 20 patients) studied: “At 99.9% sensitivity to detect ICH [intracerebral hemorrhage], the proportion of IS [ischemic stroke] patients safely differentiated was approximately 30%, whereas at 90% ICH sensitivity 65% of IS patients could be differentiated.” It’s expected that the predictive power of the algorithm will improve as larger clinical trials ramp up to provide more samples for the system to learn from.
The helmet is in early stages still, but could one day help to decrease the wait time during strokes and minimize damage to the brain by allowing earlier intervention and therapy.
Here is an animation describing the workings of the Strokefinder helmet:
Study in IEEE Transactions on Biomedical Engineering: Microwave-based stroke diagnosis making global pre-hospital thrombolytic treatment possible
Chalmers University of Technology: Strokefinder quickly differentiates bleeding strokes from clot-induced strokes