Automated in-depth cerebral arterial labelling using cerebrovascular vasculature reframing and deep neural networks

Suk-Woo Hong, Ha-Na Song, Jong-Un Choi, Hwan-Ho Cho, In-Young Baek, Ji-Eun Lee, Yoon-Chul Kim, Darda Chung, Jong-Won Chung, Oh-Young Bang, Gyeong-Moon Kim, Hyun-Jin Park, David S. Liebeskind, Woo-Keun Seo

Scientific reports   13, -(2023)

www.nature.com/articles/s41598-023-30234-6#Ack1

Clinical cases of the control, stroke and ICAS groups compared with predictively modeled vessel labels. Controls, stroke patients, and stroke-with-ICAS patients underwent colour labelling avisualisationion according to the branch within the cerebrovasculature. The algorithm consistently showed good performance for cerebral arterial branch identification among healthy controls, stroke patients, and stroke patients with ICAS. The structural abnormalities in patients with ICAS produced no significant adverse effects on identification performance.