Back to Search Start Over

Interactive exploration of a 3D intracranial aneurysm wall model extracted from histologic slices.

Authors :
Niemann, Annika
Weigand, Simon
Hoffmann, Thomas
Skalej, Martin
Tulamo, Riikka
Preim, Bernhard
Saalfeld, Sylvia
Source :
International Journal of Computer Assisted Radiology & Surgery; Jan2020, Vol. 15 Issue 1, p99-107, 9p
Publication Year :
2020

Abstract

Purpose: Currently no detailed in vivo imaging of the intracranial vessel wall exists. Ex vivo histologic images can provide information about the intracranial aneurysm (IA) wall composition that is useful for the understanding of IA development and rupture risk. For a 3D analysis, the 2D histologic slices must be incorporated in a 3D model which can be used for a spatial evaluation of the IA's morphology, including analysis of the IA neck. Methods: In 2D images of histologic slices, different wall layers were manually segmented and a 3D model was generated. The nuclei were automatically detected and classified as round or elongated, and a neural network-based wall type classification was performed. The information was combined in a software prototype visualization providing a unique view of the wall characteristics of an IA and allowing interactive exploration. Furthermore, the heterogeneity (as variance of the wall thickness) of the wall was evaluated. Result: A 3D model correctly representing the histologic data was reconstructed. The visualization integrating wall information was perceived as useful by a medical expert. The classification produces a plausible result. Conclusion: The usage of histologic images allows to create a 3D model with new information about the aneurysm wall. The model provides information about the wall thickness, its heterogeneity and, when performed on cadaveric samples, includes information about the transition between IA neck and sac. [ABSTRACT FROM AUTHOR]

Details

Language :
English
ISSN :
18616410
Volume :
15
Issue :
1
Database :
Complementary Index
Journal :
International Journal of Computer Assisted Radiology & Surgery
Publication Type :
Academic Journal
Accession number :
141098794
Full Text :
https://doi.org/10.1007/s11548-019-02083-0