Segmenting and tracking the left ventricle by learning the dynamics in cardiac images.

Having accurate left ventricle (LV) segmentations across a cardiac cycle provides useful quantitative (e.g. ejection fraction) and qualitative information for diagnosis of certain heart conditions. Existing LV segmentation techniques are founded mostly upon algorithms for segmenting static images. In order to exploit the dynamic structure of the heart in a principled manner, we approach the problem of LV segmentation as a recursive estimation problem. In our framework, LV boundaries constitute the dynamic system state to be estimated, and a sequence of observed cardiac images constitute the data. By formulating the problem as one of state estimation, the segmentation at each particular time is based not only on the data observed at that instant, but also on predictions based on past segmentations. This requires a dynamical system model of the LV, which we propose to learn from training data through an information-theoretic approach. To incorporate the learned dynamic model into our segmentation framework and obtain predictions, we use ideas from particle filtering. Our framework uses a curve evolution method to combine such predictions with the observed images to estimate the LV boundaries at each time. We demonstrate the effectiveness of the proposed approach on a large set of cardiac images. We observe that our approach provides more accurate segmentations than those from static image segmentation techniques, especially when the observed data are of limited quality.

The power-velocity integral at the vena contracta: A new method for direct quantification of regurgitant volume flow.

BACKGROUND: Noninvasive quantification of regurgitation is limited because Doppler measures velocity, not flow. Because backscattered Doppler power is proportional to sonified blood volume, power times velocity should be proportional to flow rate. Early studies, however, suggested that this held only for laminar flow, not for regurgitant jets, in which turbulence and fluid entrainment augment scatter. We therefore hypothesized that this Doppler power principle can be applied at the proximal vena contracta, where flow is laminar before entrainment, so that the power-times-velocity integral should vary linearly with flow rate and its time integral with stroke volume (SV). METHODS AND RESULTS: This was tested in vitro with steady and pulsatile flow through 0.07- to 0.8-cm(2) orifices and in 36 hemodynamic stages in vivo, replacing the left atrium with a rigid chamber and column for direct visual recording of mitral regurgitant SV (MRSV). In 12 patients, MRSV was compared with MRI mitral inflow minus aortic outflow and in 11 patients with 3D echo left ventricular ejection volume-Doppler aortic forward SV. Vena contracta power in the narrow high-velocity spectrum from a broad measuring beam was calibrated against that from a narrow reference beam of known area. Calculated and actual flow rates and SV correlated well in vitro (r=0.99, 0.99; error=-1.6+/-2.5 mL/s, -2. 4+/-2.9 mL), in vivo (MRSV: r=0.98, error=0.04+/-0.87 mL), and in patients (MRSV: r=0.98, error=-2.8+/-4.5 mL). CONCLUSIONS: The power-velocity integral at the vena contracta provides an accurate direct measurement of regurgitant flow, overcoming the limitations of existing Doppler techniques.

Relation between coronary "steal" and contractile function at rest in collateral-dependent myocardium of humans with ischemic heart disease.

BACKGROUND: We tested the hypothesis that rest asynergy in collateral-dependent myocardium correlates with coronary steal. METHODS AND RESULTS: PET with [13N]ammonia measured myocardial blood flow and flow reserve in 15 patients with symptomatic chronic ischemic heart disease. Coronary angiography assessed stenosis severity and collateral blood supply. Echocardiography or contrast ventriculography evaluated regional wall motion. Collateral-dependent segments with normal flow at rest and supplied by coronary vessels having /=0.15 mL. min-1. g-1 versus rest. Blood flow at rest in asynergic, collateral-dependent segments with steal (1.15+/-0.35 mL. min-1. g-1) exceeded (P<0.0001) that of asynergic segments without steal (0.81+/-0.24) and those with normal contraction (0.77+/-0.18). Although the flow reserve ratio of segments with normal contraction (1.8+/-0.8) exceeded that of asynergic ones with (0.6+/-0.1) or without (1.3+/-0.4) steal, overlap was great. Correlation between basal contraction and flow reserve ratio in collateral-dependent myocardium was significant but weak (r=0.45, P<0.001). However, segments demonstrating "steal" with adenosine manifested asynergy in 22 of 23 collateral-dependent segments versus 24 of 39 nonsteal segments (chi2=7.10, P<0.01). CONCLUSIONS: Although myocardial flow reserve in collateral-dependent segments with normal contraction exceeded that of asynergic segments, overlap was great. However, in patients with angina or congestive heart failure, left ventricular segments demonstrating steal with adenosine almost always exhibit asynergy at rest. Thus, coronary steal may play an important role in the pathogenesis of chronic contractile impairment at rest, whereas simple reduction of flow reserve may be less important in selected patients.

Angiotensin-converting enzyme inhibition as antiatherosclerotic therapy: no answer yet. QUIET Investigators. QUinapril Ischemic Event Trial.

Angiotensin-converting enzyme inhibitors have proven to be of clinical benefit in congestive heart failure. Whether they also provide benefit to patients with coronary artery disease in the absence of congestive heart failure via an antiatherosclerotic mechanism is a question the QUinapril Ischemic Event Trial quantitative coronary angiography (QCA) study attempted to answer: 1,750 patients with normal left ventricular function who were undergoing coronary angiography and angioplasty were randomized to 20 mg/day of quinapril versus placebo and followed for 3 years for cardiac end points. A randomly selected subgroup of the total cohort underwent follow-up angiography. The primary QCA end point was the categorical designation of progression versus nonprogression, defined either by QCA or by a cardiac event in patients selected for the QCA trial who had no usable follow-up x-ray film. Secondary end points in patients with 2 angiograms were: new stenosis development, change in minimum lumen diameter index, and change in percent diameter stenosis index. There were 119 progressors among 243 placebo-treated patients (49%) and 111 progressors among 234 quinapril-treated patients (47%) (p = NS). There were 44 patients with new stenosis development in the placebo group (19%) and 50 (22%) in the quinapril group (p = NS). Change in minimum lumen diameter index was -0.21+/-0.03 mm in the placebo group and -0.18+/-0.03 mm in the quinapril group (p = NS). Finally, change in percent diameter stenosis index was +5.1+/-1.0 in the placebo group and +3.5+/-1.0 in the quinapril group (p = NS). Potential confounders of this trial are presented and discussed.

Baseline clinical and angiographic data in the Quinapril Ischemic Event (QUIET) Trial.

The QUinapril Ischemic Event Trial (QUIET) is the first prospective, double-blind, placebo-controlled trial to investigate the long-term antiatherosclerotic effects of angiotensin-converting enzyme inhibition. Normotensive, nonhyperlipidemic subjects (1,750) with normal left ventricular systolic function were randomly assigned to treatment or placebo at percutaneous transluminal coronary angioplasty (PTCA). The primary end point is time to first cardiac ischemic event. Baseline clinical characteristics are (mean +/- SD): age 58 +/- 9 years; blood pressure 123 +/- 15/74 +/- 10 mm Hg; low density lipoprotein cholesterol 124 +/- 27 mg/dL; high density lipoprotein cholesterol 37 +/- 10 mg/dL; and triglycerides 167 +/- 91 mg/dL. In addition, 81% are men; 22% are current smokers; 49% give a history of myocardial infarction. Baseline angiographic characteristics are (mean +/- SD): left ventricular ejection fraction 59% +/- 11%; per patient diameter stenosis (excluding the PTCA segment) 49% +/- 31%; 8.9 +/- 3.5 analyzable segments per patient (excluding the PTCA segment), 3.8 +/- 2.3 of which have visible stenosis. Including the PTCA segment, 52% have single vessel disease and 48% have multivessel disease. Baseline angiographic data for non-PTCA segments will be correlated with cardiac ischemic events which occur after 6 months. Up to 500 subjects will undergo follow-up angiography with quantitative coronary angiographic analysis (QCA) of baseline and follow-up films. The primary QCA end point will be per-patient categorical designation as progressor or nonprogressor based on the presence or absence of > or = 400 microns narrowing in > or = 1 vessels that did not undergo PTCA.

Imaging and sizing of atrial septal defects by magnetic resonance.

BACKGROUND: Development of techniques for percutaneous closure of atrial septal defects (ASDs) makes accurate noninvasive sizing of ASDs important for appropriate patient selection. METHODS AND RESULTS: Magnetic resonance (MR) images of ASDs were obtained in 30 patients (mean age, 41 +/- 16 years) by both spin-echo and phase-contrast cine MR imaging. Spin-echo images were obtained in two orthogonal views (short-axis and four-chamber) perpendicular to the plane of the ASD. Spin-echo major and minor diameters were measured, and spin-echo defect area was calculated. Phase-contrast cine MR images were obtained in the plane of the ASD, and cine major diameter and defect area were measured from the region of signal enhancement or phase change due to shunt flow across the defect. MR measurements were compared with templates cut during surgery to match the defect or with ASD diameter determined by balloon sizing at catheterization. ASD size measured from cine MR images (y) agreed closely with catheterization and template standards (x). For major diameter, y = 0.78x + 5.7, r = .93, and SEE = 3.4 mm. On average, spin-echo measurements overestimated major diameter and area of secundum ASDs by 48% and 125%, respectively. CONCLUSIONS: Phase-contrast cine MR images acquired in the plane of an ASD define the defect shape by the cross section of the shunt flow stream and allow noninvasive determination of defect size with sufficient accuracy to permit stratification of patients to closure of the defect by catheter-based techniques versus surgery. Spin-echo images, on the other hand, are not adequate for defining ASD size, because septal thinning adjacent to a secundum ASD may appear to be part of the defect.

Differentiation of slow flow from thrombus in thoracic magnetic resonance imaging, emphasizing phase images.

Magnetic resonance (MR) examinations are used with increasing frequency to augment and replace other imaging modalities when a cardiovascular abnormality or thrombosis is suspected clinically. Because the signal intensity of flowing blood can vary greatly and can suggest the diagnosis of an intravascular mass, different pulse sequences and other imaging strategies may be needed to differentiate between flowing blood and intravascular thrombus. Some techniques image flowing blood as a signal void (black blood), whereas other methods enhance its signal intensity (white blood). Phase reconstruction, MR cine angiography, and related techniques that are sensitive to flow are assuming more importance in clinical MR diagnosis and are now replacing cardiac and thoracic angiography in selected clinical situations.

MRI detection of myocardial perfusion changes by gadolinium-DTPA infusion during dipyridamole hyperemia.

To detect abnormal regional myocardial coronary flow reserve, serial spin-echo MR tomograms of four control dogs and six with a partial balloon occlusion of the left anterior descending (LAD) coronary artery were acquired before and after dipyridamole infusion and during the paramagnetic effect of a bolus plus infusion of gadolinium-DTPA. Microsphere myocardial blood flow was measured for correlation with serial regional changes in MR signal intensity. Significant difference in gadolinium-enhanced MR signal intensity existed between the hypoperfused LAD and the hyperemic left circumflex perfusion beds (46 +/- 22% vs 78 +/- 29% above baseline, respectively; P less than 0.05). The percentage changes in gadolinium-enhanced MR signal and microsphere myocardial blood flow were significantly correlated (r2 = 0.93).

Noninvasive determination of valve area in adults with aortic stenosis using Doppler echocardiography.

Doppler ultrasound has been used to determine the pressure gradient P1-P2 across the valve in patients with aortic stenosis (AS), but since the gradient varies over time and may be deceptively low in patients with impaired cardiac output, the key parameter to obtain is the orifice area (A). By calculating stroke volume (SV) from the modal flow velocity [Vmode(t)] over the systolic ejection period (sep) or diastolic filling period (dfp), wherever laminar flow exists in the heart across an area of known diameter D, (pulmonary artery or atrioventricular valves), and by substituting P1-P2 = 4Vmax2, (Vmax = peak velocity in the aortic jet), the Gorlin formula becomes: (Formula: see text) where theta = flow intercept angle at D. This approach was applied in nine adult patients with AS (age 64 +/- 8 years) in whom recent catheterization data was available for comparison. Close correlation was found between the calculated areas: A(Doppler) = 0.82 A(Cath) + 0.17 (r = 0.94, p less than 0.001). Two patients with Doppler gradients of less than 40 mmHg were shown by this Doppler method nevertheless to have severely narrowed orifice areas of less than or equal to 0.78 cm2. Although there is a tendency to overestimate slightly the valve area, Doppler ultrasound assessment using this technique adds valuable noninvasive information concerning the degree of aortic valve disease.