TY - GEN
T1 - An automated segmentation and classification framework for CT-based myocardial perfusion imaging for detecting myocardial perfusion defect
AU - Qian, Zhen
AU - Joshi, Parag
AU - Rinehart, Sarah
AU - Voros, Szilard
PY - 2011
Y1 - 2011
N2 - Thanks to the recent development of the high-resolution and high-speed multi-sliced CT, CT-based perfusion imaging has become possible. In this paper, we have developed a 320-MDCT-based perfusion imaging framework to detect myocardial ischemia. We designed a rest/stress perfusion imaging protocol, developed an automated LV segmentation algorithm, and adapted a LDA-based classifier to predict myocardial ischemia using the intensity profiles in rest perfusion images. Experiments were done on 6 stress/rest CT perfusion data sets from patients with obstructive coronary artery disease (CAD) and 6 rest CT perfusion data sets from normal subjects. Experimental results have shown that rest perfusion images have the potential of accurately predicting ischemia caused by obstructive CAD.
AB - Thanks to the recent development of the high-resolution and high-speed multi-sliced CT, CT-based perfusion imaging has become possible. In this paper, we have developed a 320-MDCT-based perfusion imaging framework to detect myocardial ischemia. We designed a rest/stress perfusion imaging protocol, developed an automated LV segmentation algorithm, and adapted a LDA-based classifier to predict myocardial ischemia using the intensity profiles in rest perfusion images. Experiments were done on 6 stress/rest CT perfusion data sets from patients with obstructive coronary artery disease (CAD) and 6 rest CT perfusion data sets from normal subjects. Experimental results have shown that rest perfusion images have the potential of accurately predicting ischemia caused by obstructive CAD.
UR - http://www.scopus.com/inward/record.url?scp=79957668353&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=79957668353&partnerID=8YFLogxK
U2 - 10.1007/978-3-642-21028-0_26
DO - 10.1007/978-3-642-21028-0_26
M3 - Conference contribution
AN - SCOPUS:79957668353
SN - 9783642210273
T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
SP - 206
EP - 214
BT - Functional Imaging and Modeling of the Heart - 6th International Conference, FIMH 2011, Proceedings
T2 - 6th International Conference on Functional Imaging and Modeling of the Heart, FIMH 2011
Y2 - 25 May 2011 through 27 May 2011
ER -