Computing Publications

Publications Home » Unsupervised classification of la...

Unsupervised classification of late gadolinium enhancement cardiovascular MR

Marcus Ellington, Xiaopeng Hu, Anna John, D.J. Pennell, Guang-Zhong Yang

Conference or Workshop Paper
Medical Image Understanding and Analysis (MIUA04)

Patients suffering from ischemic left ventricular dysfunction but with viable myocardium may recover contractile function after revascularisation. The accurate identification of viable myocardium is important for guiding the subsequent treatment and establishing its prognosis. Late enhancement CMR with Gd-DTPA can depict necrotic or fibrosed areas after myocardial infarction, permitting discrimination between subendocardial and transmural defects [1]. This paper presents an automatic method for segmenting infarcted and non-infarcted regions with 3D multi-slice Late gadolinium enhancement CMR. Validation of the method was performed on data acquired from 10 patients, showing similar delineation accuracy to that of an experienced observer, but having the advantage of no inter- and intra-observer variability.

BibTEX file for the publication built & maintained by Ashok Argent-Katwala.