3D High-Resolution Cardiac Segmentation Reconstruction From 2D Views Using Conditional Variational Autoencoders

Accurate segmentation of heart structures imaged by cardiac MR is key for the quantitative analysis of pathology. High-resolution 3D MR sequences enable whole-heart structural imaging but are time-consuming, expensive to acquire and they often require long breath holds that are not suitable for pati...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Veröffentlicht in:Proceedings (International Symposium on Biomedical Imaging) S. 1643 - 1646
Hauptverfasser: Biffi, Carlo, Cerrolaza, Juan J., Tarroni, Giacomo, de Marvao, Antonio, Cook, Stuart A., O'Regan, Declan P., Rueckert, Daniel
Format: Tagungsbericht
Sprache:Englisch
Veröffentlicht: IEEE 01.04.2019
Schlagworte:
3D Segmentation Reconstruction
Cardiac MR
Deep Learning
Image reconstruction
Image resolution
Image segmentation
Solid modeling
Three-dimensional displays
Training
Two dimensional displays
Variational Autoencoder
ISSN:1945-8452
Online-Zugang:Volltext
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
Beschreibung
Zusammenfassung:Accurate segmentation of heart structures imaged by cardiac MR is key for the quantitative analysis of pathology. High-resolution 3D MR sequences enable whole-heart structural imaging but are time-consuming, expensive to acquire and they often require long breath holds that are not suitable for patients. Consequently, multiplanar breath-hold 2D cines sequences are standard practice but are disadvantaged by lack of whole-heart coverage and low through-plane resolution. To address this, we propose a conditional variational autoencoder architecture able to learn a generative model of 3D high-resolution left ventricular (LV) segmentations which is conditioned on three 2D LV segmentations of one short-axis and two long-axis images. By only employing these three 2D segmentations, our model can efficiently reconstruct the 3D high-resolution LV segmentation of a subject. When evaluated on 400 unseen healthy volunteers, our model yielded an average Dice score of 87. 92 \pm 0.15 and outperformed competing architectures (TL-net, Dice score =82.60\pm 0.23, p=2.2\cdot 10^{-16}).
ISSN:1945-8452
DOI:10.1109/ISBI.2019.8759328