Strokes are terrifying. One minute you’re walking around fine, the next minute some part of your brain is asphyxiating because a microscopic blood clot is lodged in a blood vessel. Every second counts in minimizing the damage, so doctors need to know where the clot is and what type it is as quickly as possible. Computed tomography (CT) scans are the workhorse imaging tool in the hospital, but they offer limited detail. Research-grade imaging techniques such as micro-computed tomography (the nose-shaped image on a black background) and scanning electron microscopy (the image with red blobs and a blue mesh) can reveal almost everything about a clot (colors were added digitally to distinguish tissue types), but they only work on samples that have been taken out of the body and prepared—by then the images are of no use to the patient or doctor. To get around this dichotomy, researchers at EMPA are assembling a large bank of data from clinical and laboratory instruments, and then developing machine learning algorithms to correlate clot-type information between the methods. It’s early days for the effort, but the team hopes the work will eventually help doctors pick the right interventions for stroke cases by telling them exactly what types of clots they’re dealing with.
Credit: Sci. Rep. 2022, DOI: 10.1038/s41598-022-06623-8
Do science. Take pictures. Win money. Enter our photo contest here.