PuSH - Publikationsserver des Helmholtz Zentrums München

Todorov, M.I.* ; Paetzold, J.C.* ; Schoppe, O.* ; Tetteh, G.* ; Efremov, V.* ; Völgyi, K.* ; Düring, M.* ; Dichgans, M.* ; Piraud, M.* ; Menze, B.* ; Ertürk, A.

Automated analysis of whole brainvasculature using machine learning.

bioRxiv, accepted (2019)
Tissue clearing methods enable imaging of intactbiological specimens without sectioning. Howev-er, reliable and scalable analysis of such largeimaging data in 3D remains a challenge. Towardsthis goal, we developed a deep learning-basedframework to quantify and analyze the brain vas-culature, named Vessel Segmentation & AnalysisPipeline ( VesSAP). Our pipeline uses a fully con-volutional network with a transfer learning a p-proach for segmentation. We systematically ana-lyzed vascular features of the whole brains i n-cluding their length, bifurcation points and radiusat the micrometer scale by registering them to the Allen mouse brain atlas. We reported the firstevidence of secondary intracranial collateral vas-cularization in CD1-Elite mice and found reducedvascularization in the brainstem as compared to the cerebrum. VesSAP thus enables unbiasedand scalable quantifications for the angioarchi-tecture of the cleared intact mouse brain andyields new biological insights related to the vas-cular brain function.
Weitere Metriken?
[➜Einloggen]
Zusatzinfos bearbeiten [➜Einloggen]
Publikationstyp Artikel: Journalartikel
Dokumenttyp Wissenschaftlicher Artikel
Zeitschrift bioRxiv
Verlag Cold Spring Harbor Laboratory Press
Verlagsort Cold Spring Harbor
Begutachtungsstatus Peer reviewed
Institut(e) Institute for Intelligent Biotechnologies (IBIO)