Dies ist eine Übersichtsseite mit Metadaten zu dieser wissenschaftlichen Arbeit. Der vollständige Artikel ist beim Verlag verfügbar.
EP37 Non-invasive follow-up of aneurysms treated with WEB® – A flow model analysis of CTA- and MRI-techniques versus DSA as gold standard with partially occluded aneurysms
0
Zitationen
4
Autoren
2021
Jahr
Abstract
<h3>Introduction</h3> Brain aneurysm treatment with the Woven Endo Bridge (WEB®) is widely accepted. For long term follow up a non-invasive imaging would be preferable to avoid potential risks from repetitive catheter angiography. <h3>Aim of the Study</h3> To evaluate which non-invasive imaging modality correlates best with DSA. <h3>Methods</h3> Four different realistic aneurysm models were designed and 3D printed and a WEB Device was implanted following the official sizing recommendations. Subsequently the devices were partially filled with silicone though a small borehole, intentionally leaving out a proportion of the devices volume. DSA images were made as reference, followed by MRI using T1-, T2- and TOF-sequences, as well as CT and Spectral CT scans. All Images were blinded and reviewed by two experienced readers using the WEB Occlusion Scale (WOS). <h3>Results</h3> In our model CT and Spectral CT Scans were all scored as WOS 0, resulting in 0% conformity with DSA. The readers agreed in 100% of the cases. Comparing MRI sequences with DSA consistent results were found in 9.4% of the cases. The readers scored concurring WOS in 68.8%. <h3>Conclusions</h3> From our analysis the detection of small residual inflows into aneurysms treated with WEB Devices, using non-invasive MRI or CT techniques is very unreliable and can not replace DSA for follow up. This is probably not true for recurrences outside the devices. <h3>Disclosure</h3> FW consults and proctors for Microvention. GK, MP and JM have no conflicts of interest.
Ähnliche Arbeiten
Explainable Artificial Intelligence (XAI): Concepts, taxonomies, opportunities and challenges toward responsible AI
2019 · 8.336 Zit.
Stop explaining black box machine learning models for high stakes decisions and use interpretable models instead
2019 · 8.207 Zit.
High-performance medicine: the convergence of human and artificial intelligence
2018 · 7.607 Zit.
Proceedings of the 19th International Joint Conference on Artificial Intelligence
2005 · 5.776 Zit.
Peeking Inside the Black-Box: A Survey on Explainable Artificial Intelligence (XAI)
2018 · 5.476 Zit.