HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Full Text Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Dey, D. Articles by Berman, D. S. Search for Related Content PubMed PubMed Citation Articles by Dey, D. Articles by Berman, D. S. Social Bookmarking                      What's this?

Computer-Aided Detection and Evaluation of Lipid-Rich Plaque on Noncontrast Cardiac CT

¹ Department of Imaging, Cedars-Sinai Medical Center, 8700 Beverly Blvd., Rm. 1258, Los Angeles, CA 90048.
² Tennessee Heart and Vascular Institute and EBT Research Foundation, Nashville, TN.
³ Department of Medicine, Division of Cardiology, University of California, Los Angeles, CA.
⁴ Heart Disease Prevention Program, University of California, Irvine, CA.

View larger version (14K): [in a new window]   Fig. 1 —Image histogram analysis: schematic diagram. Peak of normal blood pool image histogram is typically about 40-50 H. Coronary artery image histograms can extend below 0 H into lipid attenuation values. Gaussian curve is fitted to normal blood pool image histogram. Lipid threshold, shown by vertical line, is mean intersection value in Hounsfield units of coronary artery image histograms with normal blood pool gaussian curve. This calculated lipid threshold is specific to each cardiac CT scan.

View larger version (58K): [in a new window]   Fig. 2 —Figure shows regions of interest (ROIs) drawn and image histogram for a healthy 55-year-old man from low-risk group. Normal blood pool (BP) ROI and histogram are shown in blue; ROI and histogram for left main (LM) artery are shown in green. Fitted gaussian curve is shown in black. LM artery histogram does not intersect fitted gaussian curve; therefore, no lipid threshold or lipid-rich lesions are found.

View larger version (86K): [in a new window]   Fig. 3 —Figure shows regions of interest (ROIs) drawn and image histogram for 71-year-old man from calcium group (coronary calcium score, 11.6). Image a, which is from noncontrast electron beam tomography slice, shows proximal left anterior descending (LAD) artery. Image b shows ROIs drawn to identify normal blood pool (BP) and LAD artery. Normal blood pool histogram is shown in blue and histogram for LAD artery is shown in green. Fitted gaussian curve is shown in black. Lipid threshold is marked by red line (-0.9 H for this patient). Lipid-rich lesion found by our automated software (Plaquant) is shown overlaid in red on image b. QUANT = Plaquant.

View larger version (139K): [in a new window]   Fig. 4 —Case example 1: 64-year-old man with history of hyperlipidemia and atypical chest pain underwent electron beam tomography (EBT) coronary calcium scanning and exercise myocardial perfusion SPECT on same day. Myocardial perfusion SPECT showed large reversible anterior apical-septal defect indicating presence of hemodynamically significant lesion in proximal left anterior descending (LAD) artery. He was admitted immediately for coronary angiography. EBT scans showed minimal LAD artery calcification in next inferior slice than displayed in image a. Coronary calcium score was 8. EBT scans were, however, suspected of showing large hypodense plaque in proximal LAD artery. Subsequent quantitative analysis of lipid-rich plaque revealed lipid-laden plaque (red on EBT scans; arrow on coronary angiography and intravascular sonography images) in LAD artery beginning at origin of large first diagonal branch (images a and b) and extending for approximately 2 cm (image c). Images a and b are axial EBT views distal to origin of first diagonal artery. Image c is vertical long-axis view showing fused EBT and stress perfusion SPECT. Coronary angiography (image d) shows 80% stenosis (arrow) of proximal LAD artery at and distal to origin of first diagonal branch. Intravascular sonography confirmed presence of large plaque burden (images e and f) and 80% stenosis in proximal LAD artery. In image f, plaque on intravascular sonography is outlined in yellow and lipid-rich plaque is shown with arrow. Patient underwent stenting after intravascular sonography. QUANT = Plaquant.

View larger version (87K): [in a new window]   Fig. 5 —Case example 2: 60-year-old man with atypical chest pain underwent electron beam tomography (EBT) coronary calcium scanning and exercise myocardial perfusion SPECT. Coronary calcium score was 20 in left circumflex and right coronary arteries. No calcifications could be seen in left anterior descending (LAD) artery. Exercise myocardial perfusion SPECT showed large reversible defect indicating presence of hemodynamically significant LAD lesion. In image a (CT original), arrow shows hypodense area in LAD artery proximal to second diagonal branch, which may be lipid-rich plaque. In image b, Plaquant identified lipid-rich lesions (red) at this location. Lipid threshold for this scan was -5 H. Coronary angiography (ANGIO) (image c) performed within 1 month of images a and b revealed 80% stenosis of LAD artery (arrow) proximal to second diagonal branch. This patient underwent stenting. QUANT = Plaquant.

View larger version (23K): [in a new window]   Fig. 6 —Graph shows lipid density for coronary arteries in all three patient groups: low-risk, high-risk, and calcium groups. Asterisk indicates p < 0.05.

View larger version (21K): [in a new window]   Fig. 7 —Graph shows lipid inhomogeneity for coronary arteries in all three patient groups: low-risk, high-risk, and calcium groups. Asterisk indicates p < 0.05.

CiteULike

Complore

Connotea

Del.icio.us

Digg

Reddit

Technorati