Meeting Banner
Abstract #3750

Automated Recognition of Abnormal Left Ventricle Wall Motion

YingLi Lu1, Perry Radau1, Kim A. Connelly1,2, Alexander Dick3, Graham A. Wright1

1Imaging Research, Sunnybrook Health Sciences Centre, Toronto, ON, Canada; 2Cardiology, St Michael's Hospital, Toronto, ON, Canada; 3Cardiology, Sunnybrook Health Sciences Centre, Toronto, ON, Canada

We propose an algorithm for automated detection of hypokinetic cardiac wall motion from cine cardiac MR that includes inter-subject normalization and a pattern recognition technique. The recognition algorithm consists of three stages: 1) normalizing the left ventricle (LV) size, shape, intensity, and position, 2) extracting features called intra-segment correlation coefficients from the normalized LV images, and 3) discriminating normal and hypokinetic wall motion. Application of the algorithm on 17 patient datasets resulted in accuracy, sensitivity and specificity of 83.3%, 93.6% and 78.9% respectively. These preliminary results demonstrated a promising method for automated recognition of hypokinetic LV wall motion.