作者: Christos P. Loizou , Andrew Nicolaides , Efthyvoulos Kyriacou , Niki Georghiou , Maura Griffin
DOI: 10.1109/JTEHM.2015.2450735
关键词:
摘要: The intima-media thickness (IMT) of the common carotid artery (CCA) is an established indicator cardiovascular disease (CVD). There have been reports about difference between left and right sides CCA IMT their relation with CVD. In this paper, we propose automated system based on image normalization, speckle reduction filtering, snakes segmentation, for segmenting CCA, perform measurements, provide differences sides. study was performed 1104 longitudinal-section ultrasound images acquired from 568 men 536 women out which 125 had symptoms A expert manually delineated normal CVD groups. corresponding (normal versus CVD) mean ± standard deviation values were 0.74 0.24 0.87 mm 0.70 0.17 0.80 0.18 mm, respectively. main findings paper can be summarized as follows: 1) there no significant side IMT. These suggest that measurement one only needed group (and in agreement other studies); 2) statistical measurements both sides; 3) increasing linear relationship age group; to a lesser extend 4) found manual 5) best result classification modeling, using support vector machines, discriminate groups 64%±3.5% correct classifications score when measurements. Further research required estimating similarities intima media complex structure morphology variability texture features differentiating group.