Automatic pericardium segmentation and quantification of epicardial fat from computed tomography angiography
Artikel i vetenskaplig tidskrift, 2016

Recent findings indicate a strong correlation between the risk of future heart disease and the volume ofadipose tissue inside of the pericardium. So far, large-scale studies have been hindered by the fact that manual delin-eation of the pericardium is extremely time-consuming and that existing methods for automatic delineation strugglewith accuracy. In this paper, an efficient and fully automatic approach to pericardium segmentation and epicardial fatvolume estimation is presented, based on a variant of multi-atlas segmentation for spatial initialization and a randomforest classifier for accurate pericardium detection. Experimental validation on a set of 30 manually delineated Com-puter Tomography Angiography (CTA) volumes shows a significant improvement on state-of-the-art in terms of EFVestimation (mean absolute epicardial fat volume difference: 3.8 ml (4.7%), Pearson correlation: 0.99) with run-timessuitable for large-scale studies (52 s). Further, the results compare favorably to inter-observer variability measured on10 volumes.


machine learning

epicardial fat quantification


computed tomography angiography (CTA)


Alexander Norlén

Chalmers, Signaler och system, Signalbehandling och medicinsk teknik

Jennifer Alvén

Chalmers, Signaler och system, Signalbehandling och medicinsk teknik

David Molnar

Göteborgs universitet

Olof Enqvist

Göteborgs universitet

Rauni Rossi-Norrlund

Göteborgs universitet

John Brandberg

Göteborgs universitet

Göran Bergström

Göteborgs universitet

Fredrik Kahl

Chalmers, Signaler och system, Signalbehandling och medicinsk teknik

Journal of Medical Imaging

2329-4310 (eISSN)

Vol. 3 3 Article number 034003-


Informations- och kommunikationsteknik

Livsvetenskaper och teknik (2010-2018)



Datorseende och robotik (autonoma system)

Medicinsk bildbehandling



Mer information