Artificial intelligence (AI) accurately diagnosed 79 percent of proximal large vessel occlusions with an ipsiversive gaze deviation on non-contrast computed tomography (CT), according to new research findings presented recently at the Society of Neurointerventional Surgery’s (SNIS) 19th Annual Meeting in Toronto.
An emerging artificial intelligence (AI) algorithm may be beneficial in facilitating earlier detection of large vessel occlusions on non-contrast computed tomography (CT) scans and subsequent identification of stroke patients who are good candidates for a minimally invasive thrombectomy.
In a study abstract presentation at the Society of Neurointerventional Surgery’s (SNIS) 19th Annual Meeting in Toronto, researchers noted that the AI algorithm detects clinical symptoms of ipsiversive gaze deviation on non-contrast CT and was trained with 200 CT scans. In a subsequent study of 116 patients who received endovascular therapy for large vessel occlusions, the study authors found an ipsiversive gaze deviation in 71.1 percent of patients (59 out of 83 patients) with proximal occlusions and the AI algorithm had a 79 percent accuracy rate (47 out of 59 patients) in identifying ipsiversive gaze deviation.
The study authors said the AI algorithm could result in more expeditious treatment decisions for patients with acute ischemic stroke.
“Simply put, the faster we act, the better our stroke patients’ outcomes will be. Our results represent an advance that has the potential to speed up the identification of (large vessel occlusion) stroke during the triage process at the hospital,” emphasized lead study author Jason Tarpley, M.D., Ph.D, the stroke medical director at the Pacific Stroke and Aneurysm Center in Santa Monica, Ca.
Tarpley and colleagues acknowledged that ipsiversive gaze deviation is likely to show up on non-contrast CT images of M2 occlusions with less severe clinical symptoms. They found that only 42.4 percent of these patients had the ipsiversive gaze deviation. However, the researchers noted the AI algorithm had an 85.7 success rate in identifying the deviation in this patient population.
In comparison to control non-contrast CT scans, Tarpley and colleagues said the AI algorithm demonstrated respective sensitivity and specificity rates of 80.8 and 80.1 percent for ipsiversive gaze deviation.
“Ipsiversive gaze deviation on (non-contrast CT) is a good predictor of (large vessel occlusion) due to proximal vessel occlusions in (internal carotid artery) terminus and M1 occlusions,” noted Tarpley and colleagues.
AI Algorithm Comparable to Radiologists in Differentiating Small Renal Masses on CT
May 14th 2024An emerging deep learning algorithm had a lower AUC and sensitivity than urological radiologists for differentiating between small renal masses on computed tomography (CT) scans but had a 21 percent higher sensitivity rate than non-urological radiologists, according to new research.
What a New Meta-Analysis Reveals About Fractional Flow Reserve Assessment with Computed Tomography
May 13th 2024While acknowledging variable accuracy overall with CT-derived fractional flow reserve (FFR-CT) values, researchers found that the accuracy rate increased to 90 percent for FFR-CT values greater > 0.90 and < 0.49.
Study Finds High Concordance Between AI and Radiologists for Cervical Spine Fractures on CT
May 6th 2024Researchers found a 98.3 percent concordance between attending radiology reports and AI assessments for possible cervical spine fractures on CT, according to new research presented at the 2024 ARRS Annual Meeting.