AI helps clinicians ID stroke candidates for thrombectomy
A new machine learning algorithm can determine which stroke patients would benefit from an endovascular thrombectomy based off of CT angiography (CTA) scans, according to new research out of the University of Texas Health Science Center at Houston.
Thrombectomy is a procedure that allows clinicians to mechanically remove a brain clot and has been demonstrated to greatly improve outcomes for stroke patients, said Sunil A. Sheth, MD, assistant professor of neurology at UTHealth, in a prepared statement. However, many care centers don’t have access to this technology.
"With endovascular thrombectomy, we now have a treatment for ischemic stroke that is really revolutionary. It allows us to take stroke patients from severe disability and return them to an almost normal life," Sheth added. “Unfortunately, the advanced imaging techniques used currently to identify which patients benefit from this procedure are not widely available outside of large referral hospitals. As a result, most stroke patients don't have access to guideline-based screening for these treatments."
The researchers created a machine learning platform—DeepSymNet—to identify large vessel occlusion (LVO) and infarct core from CTA images. It can also be used as a proxy for other advanced imaging modalities such as CT perfusion, which are not widely available in stroke clinics or hospitals.
Sheth and colleagues tested the algorithm on 297 patients within their stroke registry who had experienced a stroke or had conditions that “mimicked” one. Of the 224 who had a stroke, 179 had blocked cerebral blood vessels or LVO. DeepSymNet learned to identify intracerebral vasculature on CTA images and detected LVO with receiver-operative curve area under the curve (AUC) of 0.88. The algorithm also determined infarct core from images with an AUC of 0.88 and 0.90. Importantly, DeepSymNet produced similar accuracies in patients with early and late time windows, the authors noted.
The findings, according to Sheth et al., show that the information required to evaluate patients for thrombectomy may be within CTA images.
"The advantage is you don't have to be at an academic health center or a tertiary care hospital to determine whether this treatment would benefit the patient,” Sheth said. “And best of all, CT angiogram is already widely used for patients with stroke.”
The full study was published online Sept. 24 in Stroke.
Matt joined Chicago’s TriMed team in 2018 covering all areas of health imaging after two years reporting on the hospital field. He holds a bachelor’s in English from UIC, and enjoys a good cup of coffee and an interesting documentary.