YEAR IN CARDIOLOGY SERIES
The Year in Echocardiography
Arthur E. Weyman, MD*
Cardiac Ultrasound Laboratory, Massachusetts General Hospital, Boston, Massachusetts
Manuscript received January 12, 2009;
accepted January 21, 2009.
* Reprint requests and correspondence: Dr. Arthur E. Weyman, Cardiac Ultrasound Laboratory, Yawkey 5E, Massachusetts General Hospital, Boston, Massachusetts 02114 (Email: aweyman{at}partners.org).
Key Words: echocardiography • year • 2008
The evolution of all noninvasive diagnostic tests ideally proceeds in orderly stages, beginning with description and initial enthusiasm, followed by reassessment and re-evaluation, and finally clinical, research, and epidemiologic application. However, for techniques such as echocardiography that are widely available, that have no inherent harmful effects, and for which the requirements for proof of efficacy are limited, clinical/research application often occurs coincident with or may precede reassessment and re-evaluation, which often awaits the development of new or improved methods of analysis (the classic example being mitral valve prolapse). As a result, one often finds periods in which publications appear almost simultaneously by proponents of a technique or application whereas others point out its limitations. Application before reassessment and refinement is a particular problem in areas in which there is no true clinical gold standard or the reference method has its own limitations (e.g., mitral regurgitation). The year 2008 in particular was characterized by reassessment and re-evaluation in many areas, including: 1) the use of echo Doppler techniques to identify patients for cardiac resynchronization therapy (CRT); 2) the role of left ventricular torsion in the assessment of diastolic function; 3) contrast echocardiography; and 4) the Doppler quantification of mitral regurgitation.
 |
CRT
|
|---|
Over the past several years, a number of echocardiographic techniques and criteria have been reported to aid in the identification of patients most likely to respond to cardiac resynchronization therapy before device implantation. These data in general arise from single-center studies, often with conflicting results, and there has been no general agreement on the ideal measurement technique, the criteria for defining dyssynchrony (DYS), or the method/criteria for defining a positive response to CRT. Two recent studies have raised additional questions about the utility of various echocardiographic measures in predicting response to CRT. In a prospective, multicenter setting, the PROSPECT (Predictors of Response to CRT) study examined 426 New York Heart Association (NYHA) functional class III or IV patients with an ejection fraction (EF)  35% and a QRS duration  130 ms (1). Echocardiographic data from 53 contributing laboratories were analyzed in 3 core laboratories (1 in the U.S., 1 in the United Kingdom, and 1 in Italy). Intraobserver and interobserver variability were 3.6% and 14.5% for left ventricular (LV) end-systolic volume (ESV), 3.7% and 6.5% for LV pre-ejection period, 24.3% and 72.1% for M-mode septal–posterior wall motion delay, 11.4% and 33.7% for the standard deviation of the interval time from the onset of the QRS to the time to peak systolic velocity (Ts-SD) for 12 segments (6 basal and 6 midventricular), and 15.8% and 31.9% for maximal difference in the time to peak systolic velocity for the 6 basal segments (Ts-peak [basal]). Two outcome measures were used: 1) a clinical composite score (CCS) that combined all-cause mortality, heart failure (HF) hospitalizations, NYHA functional class, and a patient global assessment; and 2) a reduction of LVESV 15%. Initially, the core laboratories identified 20% of the enrolled subjects as having EFs that exceeded the entry criteria. Although several parameters predicted statistically significant improvement in clinical and reverse remodeling outcomes, sensitivity and specificity were modest. The highest area under the curve for the any of the 12 methods tested was 0.60 for prediction of improvement by the CCS and 0.61 for a reduction of LVESV 15%. There was a relatively low yield and high variability in the tissue Doppler imaging (TDI) measures, with the percent of individual parameters deemed interpretable ranging between 37% and 82%. As pointed out in the accompanying editorial, there are a number of problems with using TDI for timing measurements, including the difficulty in defining peak contraction in the relatively flat velocity contour of the failing heart, signal noise, spatial and temporal variation of the signal, and the fact that TDI measures motion toward the transducer whether active or passive (rocking motion or tethering) (2). The investigators conclude that "current iterations of echocardiographic and TDI indexes of mechanical synchrony are unsuited to everyday clinical use in CRT selection. These measurements are too variable and their sensitivity is currently too low to use in the hope of avoiding nonresponders..." (1).
Small single-center studies have also reported that TDI and other related techniques identify a significant percentage of patients with HF and normal QRS intervals as having dyssynchrony and that these patients may benefit from CRT. In a prospective, randomized controlled study designed to examine the value of resynchronization therapy in patients with HF and normal (<130 ms) QRS intervals (RETHINQ [REsynchronization THerapy IN normal QRS]), 172 patients were randomized equally to CRT and control groups. Echocardiographic evidence of mechanical dyssynchrony was an entry requirement, and dyssynchrony based on an increased septal–lateral wall delay by TDI 65 ms was present in all patients in both groups. At 6 months, there was no difference between the control and CRT groups in the primary end point (an increase in peak oxygen consumption of at least 1.0 ml/kg/min), nor was there a significant difference in other end points, including the quality-of-life score, 6-min walk test, or LV reverse remodeling, although NYHA functional class improved in the CRT group. Thus, the study found no benefit of CRT in patients with a QRS interval 130 ms who met TDI criteria for dyssynchrony (3).
Although these studies seemingly question the accuracy and/or utility of echocardiographic/Doppler methods in identifying appropriate patients for CRT, it is important to remember that factors other than the ability of the diagnostic technique to accurately define dyssynchrony can be responsible for failure to respond to CRT. These include areas of LV scar that cannot contract even if appropriately paced and failure to position the pacing catheter adjacent to the area of greatest contraction delay because of the LV venous anatomy. Although scar in the area of dyssynchrony has been shown by magnetic resonance imaging (MRI) to be a cause of response failure with CRT, the demonstration of preserved function in the dysynchronous region should exclude scar and indicate the potential of the region to effectively contribute to global LV performance with appropriately timed contraction. In an attempt to assess the potential for incremental improvement in LV contractility after time alignment of delayed segments by CRT, Lim et al. (4) proposed a new strain delay index (SDI) based on longitudinal strain (LS) assessed by speckle tracking (ST) and compared it with a TDI-based index of dyssynchrony (SD of the time to peak velocity in 12 basal segments). The SDI was calculated as the sum of the difference between the peak and the end-systolic LS for 16 myocardial segments. The index seems based on the assumption that the time of peak strain in segments with synchronous contraction will be close to end systole, and thus peak strain will differ little from end-systolic strain and hence contribute little to the index. Areas of scar will not contract (approximately 0 value) and thus will not represent wasted or potential recoverable energy and likewise will not contribute to the index. As a result, the index should represent predominately the post-systolic contraction of dysynchronous but functional (as opposed to scarred) segments with higher index values corresponding to greater wasted (and presumably recoverable) energy. In a derivation group of 65 patients, receiver-operator characteristic curve data showed an optimal index cutoff of 25% to predict a positive response to CRT (>15% reduction in ESV at 3-month follow-up). In a validation group (n = 35), the SDI was significantly higher in responders (32 ± 8%) than in nonresponders (20 ± 5%), and an SDI 25% identified 82% of responders (18 of 22) and 92% of nonresponders (12 of 13). The SDI correlated with reverse remodeling (decrease in ESV) (r = –0.86, p < 0.0001 for all patients). As with the PROSPECT and RETHINQ studies, the investigators found that TDI time delay measurement failed to improve patient selection. Although conceptually intriguing, the practical application of this index may be limited because of the inability to obtain adequate images in many patients (approximately one-third of patients in the derivation group had to be excluded), interobserver variability (20% in the validation group), sensitivity (18% of responders in the validation group would have been missed based on the derived threshold), and exclusion of patients with atrial fibrillation. Nevertheless, this study reinforces the need to account for the preservation of function in the area of dyssynchrony and its potential to contribute to overall ventricular function with appropriate timing of contraction. When combined with methods to more precisely align the pacing catheter with the area of greatest energy loss, this type of measurement could provide a more robust approach to patient selection for CRT (4).
The second factor that can lead to failure of response to CRT is the inability to position the LV pacing lead adjacent to the area of greatest dyssynchrony because of the limitation of cardiac venous anatomy. In a study designed to compare the clinical and echocardiographic response to CRT in patients with concordant (LV lead placement adjacent to the area of latest mechanical activation) versus discordant (lead was placed outside this area) lead placement, Ypenburg et al. (5) followed up 257 patients for a mean of 32 months after CRT. The site of the latest mechanical activation was determined from midventricular short axis radial strain (RS) curves (ST) and was most frequently located in the lateral (33%) and posterior (36%) walls. Lead placement was concordant with the site of latest activation in 63% of patients and discordant in 37%. After 6 months of CRT, patients with concordant LV leak position showed significant improvement in echocardiographic parameters (LV end-diastolic volume, LVESV, EF). In contrast, patients with a discordant LV lead position showed no significant improvement in LV volumes or EF. Event-free survival for the primary clinical end points (death, hospitalization for HF, and transplantation) were quite similar for up to 24 months; however, longer follow-up showed worse outcome (3-year event-free survival) in patients with discordant lead placement, 57% versus 78% in the concordant group (5). Thus, to truly determine the sensitivity and specificity of echo-Doppler methods to predict response to CRT (as opposed to identifying DYS), future studies will need to account for the presence of scar and appropriateness of lead placement. Unfortunately this will not correct for measurement variability or eliminate the patients with severe HF who do not meet criteria for mechanical DYS and yet respond to CRT.
|
Speckle Tracking for CRT
|
|---|
Given the limitation of TDI in identifying responders to CRT, alternative methods have been examined. Several prior studies have suggested that the time of peak strain by ST might offer an alternative method for assessing dys-synergy, although the results have been inconsistent (6,7). To examine this question further, Delgado et al. (8) studied 161 patients with HF (NYHA functional class III or IV), EF >35%, and QRS duration >120 ms before and 6 months after CRT. Time to peak LS, RS, and circumferential strain (CS) was assessed by ST in the apical 2- and 4-chamber (LS) and the parasternal short axis (mid-papillary muscle) views (RS and CS). For each type of strain, the maximum time delay between peak systolic strain in 2 of the 6 recorded segments and the SD of the time to peak strain was calculated. The RS could be calculated by ST in 90% of segments, CS in 85%, and LS in 79%. At 6 months, 55% of patients were classified as responders (reduction in LVESV >15%). For RS, the septal–posterior wall delay (S-P) and the SD of the time to peak radial (SD-t6s) were significantly larger for responders compared with nonresponders. However, there was no significant difference in CS or LS between groups. There was a weak but significant correlation between time to peak wall delay (r = 0.41) and SD-t6s by RS (r = 0.26) and decrease in ESV. A cutoff value for radial dyssynchrony (S-P delay) of 130 ms was able to predict response to CRT with a sensitivity of 83% and a specificity of 80%, whereas the optimal cutoff for SD-t6s of 76 ms yielded a sensitivity of 77% and a specificity of 60%. A higher value of baseline radial DYS corresponded with a larger reduction in LVESV after 6 months of CRT. In addition, a significant decrease in the extent of LV dyssynchrony measured with RS was shown only with responders. In contrast to the prior studies, responders also showed more baseline LV DYS than nonresponders by TDI (maximal difference among 4 basal segments). Using a cutoff of 65 ms yielded a sensitivity and specificity for response of 81% and 63%, respectively. This study shows that evaluation of DYS using ST strain analysis is feasible and that substantial DYS is present in all 3 deformation types (LS, CS, and RS) in CRT candidates with depressed LV function and dilated cardiomyopathy. However, only baseline RS patterns were able to identify potential responders to CRT (8).
|
Torsion
|
|---|
The magnitude and rate of twisting of the ventricle as it contracts (torsion) and the subsequent untwisting during relaxation have been topics of interest for the past several years. The primary focus has been on the rate of untwisting, which several reports suggest correlates with tau ( ), the time constant of LV relaxation, and thus might offer an improved noninvasive measure of diastolic function. In an experimental study to examine the relationship of untwisting, relaxation (), and early diastolic intraventricular pressure gradients (IVPGs) (diastolic suction), Notomi et al. (9) examined the sequence of events that occurs during early diastole (encompassing isovolumic pressure decay and rapid LV filling) in the normal LV at baseline, during RV and LV pacing, and during dobutamine and esmolol infusion. They observed that peak twisting coincided with end systole, followed in sequence by the peak untwisting rate, which coincided with mitral valve opening, the peak rate of LV long axis relengthening and short axis expansion, and the peak IVPG, which occurred after both mitral valve opening and peak untwisting rate. This was followed by the peak mitral E-wave. Repeated-measures regression models showed moderately strong correlations between peak LV twisting with peak untwisting rate (r = 0.74), the time constant of LV pressure decay () (r = –0.66), and IVPG (r = 0.76, p < 0.0001 for all). Interestingly, RV pacing caused a decrease in peak torsion from 6.6° ± 3.4° to 2.7° ± 2.1° and in peak untwisting velocity 0.99 ± 0.20 rad/s to 0.47 ± 0.32 rad/s, but no significant change in end-diastolic volume, ESV, or EF (9).
In another study, Wang et al. (10) examined the relationship between the LV untwisting rate (UR) and diastolic function in a combined experimental and clinical study. In the experimental studies, LV untwisting rate (°/s) showed the best correlation with ESV (r = –0.8). At baseline, inferior vena caval occlusion led to a significant decrease in ESV and an increase in UR without a significant change in . Conversely, in experimental stages in which LVESV was constant, LV URs were similar despite large and significant differences in . Thus, although no correlations were given, there did not seem to be a significant relationship between UR and . In the clinical study, the investigators compared normal subjects with patient groups with systolic dysfunction and diastolic dysfunction with normal EFs. The LV twist and twisting rate were no different from that of control subjects in patients with isolated diastolic dysfunction but were significantly reduced in patients with systolic dysfunction. Likewise, the UR was similar in patients with isolated diastolic dysfunction and normal control subjects, but it was reduced in patients with systolic dysfunction. In patients with systolic HF, UR was related to (r = –0.78), but this was not the case in patients with isolated diastolic HF, in whom the correlation between UR and was r = –0.29, p = 0.15. Thus, although UR seems to have a consistent relationship with twist and ESV, its relationship to is much less certain (10).
The relationship between torsion and LV function has been examined in several other conditions. In a study comparing 5 patients with congenital total absence of the pericardium with 10 control subjects, there was no significant difference in EF, LS, RS, or CS or systolic and early diastolic strain rates between the 2 groups. However, LV torsion was markedly decreased in patients with pericardial defects compared with the normal group (approximately 4.5° vs. 20.5°) (11). Although the normal values for torsion in this study are quite high, the data raise further questions about the obligate linkage between LV torsion and LV function. In another recent study, Pettersen et al. (12) showed that torsion was absent in the systemic right ventricle in patients with transposition of the great vessels who underwent the Senning procedure despite preserved systolic and diastolic function.
Taken together, this combination of data from controlled experiments, clinical studies, and experiments of nature identifies a variety of circumstances in which torsion is not directly connected to or does not track changes in LV systolic or diastolic function. Thus, although torsion may improve the efficiency of LV contraction (13), the presence of torsion does not seem essential for the ventricle to function normally, nor can it be used as a measure of systolic or diastolic function in all settings. As Notomi et al. (9) point out in their conclusion, "this study demonstrates the correlations between and temporal succession of, isovolumic pressure decay, LV untwisting, and IVPG but does not address a cause and effect relationship between these phenomena."
|
Speckle Tracking
|
|---|
Accuracy in defining local function during dobutamine stress echocardiography.
Data supporting the accuracy of 2-dimensional ST for determining the magnitude and timing of strain for the assessment of resting myocardial function has prompted interest in the use of this method to quantify regional function during dobutamine stress echocardiography (DSE). However, because the frame rates of ST images used to derive strain data are relatively low (40 to 90 frames/s), there is concern that strain rates may not be accurate at the higher heart rates associated with stress (14). To evaluate the potential of ST to detect ischemia during DSE, Reant et al. (15) compared LS, CS, and RS by ST with sonomicrometer-assessed strain in an ovine model of flow limiting stenosis (FLS) and non–flow limiting (reduction in hyperemic response) stenosis (NFLS). Although a moderately good correlation was noted between sonomicrometry and ST-derived strain at rest and during dobutamine infusion for all strain measures, the slopes were relatively flat, varying from 0.80 for LS at rest to 0.38 for CS during dobutamine. At rest, FLS 25% significantly reduced LS (approximately –27%), whereas FLS 50% further decreased LS while also decreasing CS (approximately –26%). The RS was not significantly reduced at rest (by the principle of conservation of mass, it is difficult to understand how resting CS and LS could decrease in response to a 25% and 50% flow-limiting stenosis and RS fail to do so; further, the resting correlation between ST-derived RS with sonomicrometer data was only r = 0.56 with a slope of 0.48, the interobserver and intraobserver variability for RS was approximately 15%, and fewer data points could be analyzed, raising questions about the accuracy of the method to measure RS in this model). The effects of coronary stenosis on different strains were more pronounced during dobutamine infusion, NFLS 70% induced a significant reduction in LS (approximately 17%) and CS (approximately 17%) but not in RS. The FLS 50% significantly decreased LS (approximately 39%), CS (approximately 26%), RS (approximately 28%), and wall thickening (approximately 29%). Thus in this open chest model ST was able to detect directionally similar changes to those observed by sonomicrometry in LS, CS, and RS, and was sensitive to FLS both at rest and during dobutamine infusion (15).
Detection of viability and improvement in function after revascularization.
In another study, Becker et al. (16) compared ST with contrast-enhanced magnetic resonance imaging (ceMRI) for detecting viability and improvement in cardiac function after revascularization (either percutaneous transluminal coronary angioplasty or bypass surgery) in 53 patients in sinus rhythm with chronic ischemic heart disease. Peak RS and CS were determined from short-axis grayscale images at the base, mid-ventricle, and apex (16 segments) and compared with the percent hyperenhancement by ceMRI for the same segments. Functional recovery for segments (>1 grade) was determined visually by comparison of echocardiographic images, whereas global recovery was defined as an increase in EF of >5%. Image quality was adequate for echocardiographic analysis in 91% of segments. Of 463 dysfunctional segments at baseline, 227 segments (49%) showed functional recovery at follow-up (9 ± 2 months) and 236 (51%) did not. A peak systolic RS >17.2% had a sensitivity of 70.2% and a specificity of 85.1% to predict segmental functional recovery. When all initially dysfunctional segments were analyzed and quintiles of peak strain were considered, the percent showing improved contractility decreased progressively as the peak RS decreased. Thus, contractility increased after reperfusion in 82 of 93 segments (88%) with peak systolic strain >22.9% but in only 3 of 64 segments (5%) with peak systolic RS <5.5%. A similar pattern was noted for percent hyperenhancement by ceMRI with contractility increasing in 117 of 140 segments (84%) with no hyperenhancement and in only 1 of 50 segments (2%) with hyperenhancement involving >75% of tissue. Using receiver-operator characteristic–defined cutoff points of >17.2% for peak systolic RS and 43% hyperenhancement to define viability, an increasing extent of dysfunctional but viable myocardium before revascularization correlated with a greater improvement in ejection fraction after revascularization (r = 0.75 for RS, r = 0.77 for ceMRI). There were no significant differences between the 2 techniques in the prediction of functional segmental or global recovery (16).
|
TDI
|
|---|
Effects of valve closure on the longitudinal velocity profile.
The TDI myocardial velocity recordings from the cardiac apex typically contain a positive systolic shortening wave (S-wave) and negative displacements representing LV lengthening during early filling (E-wave) and atrial contraction (A-wave). In addition, there are 2 periods of rapid oscillation or spikes just before and just after the ejection wave (Fig. 1). Although several mechanisms have been proposed to explain these spikes, the exact mechanism has remained undefined. In a series of experimental studies, Remme et al. (17) showed that myocardial shortening started 20 ± 10 ms before mitral valve closure and was arrested by mitral valve closure. Similarly, myocardial lengthening started 31 ± 15 ms before aortic valve closure and was interrupted at the time of aortic valve closure. Prevention of mitral and aortic valve closure by stenting abolished the pre- and post-ejection spikes, respectively. Studies in patients with severe mitral and aortic regurgitation showed that the respective pre- and post-ejection spikes were reduced after valve replacement. The previously suggested alternative mechanisms of asynchronous onset of contraction/relaxation and atrioventricular interaction were investigated as possible causes of velocity spikes in animals and patient groups but were found implausible (17). Thus, the pre-ejection spike seems to be caused by initial LV wall shortening and displacement of blood toward the mitral valve that is arrested by mitral valve closure, whereas the post-ejection spike is similarly caused by early relaxation of the LV that was arrested by aortic valve closure.
Estimation of LV filling pressure in advanced heart failure.
Prior studies have reported that the ratio of the early transmitral filling velocity to the TDI mitral annular early diastolic velocity (E/Ea) correlates with the pulmonary capillary wedge pressure (PCWP) in a variety of conditions. To examine the relationship between E/Ea and PCWP in a group of patients with advanced decompensated heart failure (ADHF), Mullins et al. (18) studied 106 patients with LVEF 30% (mean 24 ± 8%) and NYHA functional class III or IV undergoing right heart catheterization. The investigators found no correlation between PCWP and mitral Ea in either the septal or the lateral annulus or when averaged over both annular regions. Likewise, there was no significant correlation between the mitral E/Ea ratio (using the septal, lateral, or mean Ea) and PCWP (r = 0.18, p = 0.07). The sensitivity and specificity of a mitral E/Ea ratio >15 to identify a PCWP >18 mm Hg were 66% and 50%, respectively. The predictive value was similar when patients with atrial fibrillation were excluded. However, a mitral E/Ea ratio >15 provided better accuracy in predicting PCWP >18 mm Hg in patients without prior CRT device implantation (51%) than in those with a device (sensitivity 72% vs. 59%, specificity 54% vs. 52%). To determine the potential importance of a cutoff value for E/Ea, patients were divided into those with E/Ea ratios <8, 8 to 15, and >15; however, PCWPs were similar for each group (19 ± 4 mm Hg, 19 ± 7 mm Hg, and 20 ± 6 mm Hg, respectively). Follow-up studies were obtained in 49% of patients, and there was no correlation between absolute change in PCWP and change in mitral E/Ea ratio. The investigators observed discordance between the mitral E/Ea ratio and PCWP to be linked to larger LV dimensions, more impaired cardiac output, and the presence of CRT. They conclude that their "observations underscore the need for refinement in the broad clinical use of the mitral E/Ea ratio to estimate filling pressures and caution against the direct inference of such relationships in patients in the decompensated state with significant LV systolic dysfunction, cardiac remodeling, or biventricular pacing" (18).
|
Pulsed Doppler: Restricted Filling Pattern
|
|---|
Prognosis in acute myocardial infarction (AMI).
A restrictive filling pattern (RFP) by pulsed Doppler assessment of mitral inflow (high peak E-wave, increased E/A ratio, short deceleration time) has been shown in several studies to be a predictor of poor outcome in patients with AMI. However, these studies have been limited by small sample size and low event rates, which preclude a definite conclusion as to the independent importance of an RFP when overall LV function is preserved. As a result, the Meta-analysis Research Group in Echocardiography (MeRGE) undertook an individual patient meta-analysis to determine whether RFP is predictive of mortality independent of LVEF, ESV index, and Killip class in patients after AMI (19). A total of 3,396 patients from 12 prospective studies were analyzed. Patients with RFP were characterized by a more frequent history of diabetes, prior AMI, hyperlipidemia, higher Killip class, anterior MI, larger LV volumes, and lower LVEF. During follow-up (no-RFP median = 1,060 days; RFP median = 1,072 days), 192 patients (29%) with RFP and 307 (11%) without RFP died (hazard ratio 2.67; 95% confidence interval: 2.23 to 3.20; p < 0.001). The RFP was associated with a higher all-cause mortality independent of age, sex, LVEF, ESV, or Killip class. The overall prevalence of RFP was 20%, but was highest (36%) in the quartile with the lowest LVEF (<39%) and lowest (9%) in the quartile with the highest LVEF (>53%). The 90% survival times were shortest in the lowest EF group, but within each quartile, survival time was markedly reduced when patients with restrictive filling were compared to those with nonrestrictive filling. When diabetes, current medication, and prior MI were included in the model, RFP remained an independent predictor of outcome. Thus, although patients with a low LVEF and dilated ventricle have a poor prognosis, this study suggests that an RFP helps identify higher-risk patients even within these groups as well as identifying higher-risk patients with normal LVEF (19).
Occurrence of a restrictive filling pattern as a result of volume loading with normal LV function.
Although an RFP is generally associated with impaired LV diastolic function with an elevated LA pressure, it has been questioned whether a similar pattern may occur with an increase in LA pressure simply because of rapid volume loading (RVL) of the LV. To examine this question, Masutani et al. (20) studied the LV filling patterns in conscious dogs with a similar elevation in LA pressure caused by either RVL or pacing-induced HF with RFP. The LA pressure and heart rate were similar after volume loading and HF. In both cases the peak E was higher than in normal control subjects, the E/A ratio was similarly elevated, and the deceleration slope was similarly steep. Thus, both showed RFP. Other features, however, permitted differentiation, including a shorter deceleration time in HF than RVL, an E' that was reduced in HF but increased in RVL, and a delay in E' relative to E in HF. Of note, despite similar LA pressures (mean LA pressures [mm Hg] for control = 7.1, VL = 22.1, HF = 22.6) there was a 93% mean increase in the E/E' ratio with HF compared with only a 40% increase with VL, raising questions about the value of this ratio in predicting LA pressure in the setting of normal diastolic function.
|
Contrast Echocardiography
|
|---|
Safety of contrast echocardiography.
In October 2007, the U.S. Food and Drug Administration announced a new "black box" warning for the perflutren-containing ultrasound contrast agents Optison (GE Healthcare, Buckinghamshire, United Kingdom) and Definity (Bristol-Myers Squibb Medical Imaging, Billerica, Massachusetts) stating that their use was contraindicated in patients with acute coronary syndromes, AMI, and worsening or unstable HF. These warnings followed post-marketing reports of 4 deaths that were temporally related but not clearly attributable to contrast agents because any or all could reasonably have been caused by the patients' underlying conditions. These concerns, although appropriate, would have severely limited the use of echo contrast in situations in which it might be most necessary. In addition, the contraindications seemed inconsistent with the initial safety data accumulated before product approval and the widespread and seemingly safe use of contrast (agitated saline and commercially prepared microbubbles) by the echocardiographic community. Even if all of the reported events occurred as a direct result of the contrast administration (which is unlikely given the underlying disease in these cases), the risk of death would be 1:500,000 based on reported events and doses administered since product approval. Given a mortality rate this low, it would be practically impossible to perform a prospective randomized trial to address mortality; however, the warning did prompt several studies to look at the safety of echo contrast in general. In 1 study, Kusnetzky et al. (21) compared the death rates within 24 h after echocardiography for 18,671 patients studied between January 2005 and October 2006 in the St. Luke's health system, Kansas City, Missouri. Death within 24 h for patients undergoing echocardiography without Definity (n = 12,475) was 0.37% compared with 0.42% for those receiving Definity (n = 6,196) (p = 0.60), despite evidence of higher acuity and more comorbid conditions in patients receiving contrast. There were no deaths within 1 h of contrast injection (21). In a second retrospective study echocardiography and general ultrasound laboratories known to have a high volume of contrast use in the U.S. and Canada were invited to submit data via an online questionnaire. Of the 33 sites invited to participate, 12 cardiac laboratories and 1 general ultrasound laboratory provided the requested data including the complications associated with 78,383 contrast doses (66,164 Definity and 12,219 Optison). Severe reactions that were considered probably related to an ultrasound contrast agent (Definity) occurred in 8 patients (0.01%), all of whom were outpatients, with 4 of these consistent with anaphylactoid or complement activation related pseudoallergy (CARPA) reactions. (CARPA has many features similar to immunoglobulin-E–mediated type-1 allergic reactions, including angioedema, bronchospasm, hypotension, hypoxemia, low back or lumbar pain, and urticaria. However, CARPA has some distinctive features, including the development of reactions without prior exposure to the allergen, tendency for reactions to decrease in severity or disappear over time, and the potential for spontaneous resolution [22]). All occurred with 30 min of contrast injection. There were no deaths, and all patients recovered with treatment. No serious events were seen in inpatients, including over 10,000 administrations given to acutely ill patients in intensive care settings or those with acute chest pain of suspected cardiac origin (22). Although such studies have the obvious potential for referral bias, the data are consistent with the general clinical experience and suggest that the incidence of severe adverse reactions to ultrasound contrast agents is no greater, and may be lower, than that reported for contrast agents commonly used in other cardiac imaging tests (22).
Molecular imaging.
There is growing interest in the development of methods using echocardiographic contrast to image disease at the level of its cellular and/or molecular mediators. Molecular imaging with contrast-enhanced ultrasound relies on the detection of novel site-targeted microbubble contrast agents that are retained within or bind to specific components of a disease process, thereby allowing phenotypic characterization. Most prior studies have used microbubbles targeted to a single molecular receptor to identify its presence and concentration. However, by targeting multiple receptors it might be possible to analyze more complex molecular and cellular processes. To examine the inflammatory responses that contribute to vascular remodeling during tissue repair or ischemia, Behm et al. (23) used multiple contrast agents separately targeted to endothelial vascular cell adhesion molecule-1, complement receptors expressed preferentially by activated neutrophils, and the fibronectin receptor ( 3β1) on monocytes to study the timing of endothelial cell activation and inflammatory cell recruitment during vasculogenesis and ischemia-mediated angiogenesis in models of chronic ischemia. When combined with contrast perfusion imaging, this allowed the molecular events to be temporally related to new vessel formation and increased blood flow and showed the potential for studying how vascular remodeling can be modulated for therapeutic effect (23). Although these techniques are clearly still in the early phase of development and it remains to be determined whether similar data would be obtainable in patients with less favorable windows and greater background noise from tissue and circulating bubbles, the potential of such methods is enormous (24).
Vasa vasorum imaging.
Another area of ongoing interest is the development of methods to assess the atherosclerotic plaque neovascularization (increased vasa vasorum) that seems essential for plaque growth, accompanies inflammation, and has been associated with plaque vulnerability and symptomatic disease (25,26). In a study comparing contrast echocardiographic imaging of carotid plaques with surgical endarterectomy samples, Coli et al. (25) examined 52 plaques from 32 patients. Plaques with more intense contrast enhancement showed a significantly greater degree of neovascularization by histology. In contrast, there was no correlation between the degree of stenosis and contrast enhancement. When plaques where characterized as echolucent or echogenic, a positive correlation between echolucency and more intense neovascularization was still present. Although early, the development of a technique to study human plaque neovascularization in vivo through direct visualization of the vasa vasorum may allow the assessment of the response to antiatherosclerotic therapies and improve risk stratification (25).
In another study, the extent of microvascular damage detected and quantitated by myocardial contrast echocardiography was the most powerful independent predictor of LV remodeling after ST-segment elevation myocardial infarction compared to ST-segment elevation, Thrombolysis In Myocardial Infarction flow grade, or myocardial blush grade (27).
|
Patent Foramen Ovale and Migraine
|
|---|
A number of case-control studies have noted an association between PFO and migraine and a reduction in headache frequency has been reported following percutaneous PFO closure in uncontrolled studies. To examine the association of PFO and migraine in an urban, population based, multiethnic cohort, Rundek et al. (28) studied 1,101 stroke free subjects from the ongoing NOMAS (Northern Manhattan Study) study. The prevalence of PFO and migraine (self-reported) were 15% and 16%, respectively. The prevalence of PFO was not significantly different between subjects who had migraine (26 of 178 or 14.6%) and those who did not (138 of 923 or 15%, p = 0.9). PFO was slightly more prevalent in migraine with aura (19%) than in migraine without aura (11%) but this did not reach statistical significance (p = 0.42). The lack of association was not modified by age or sex. While raising important questions about the association of PFO and migraine, the study has several limitations. First, the population is older (mean age 69 ± 10 years) so that a potential association in younger patients cannot be excluded. Second, the prevalence of PFO is lower than expected, which may reflect the use of transthoracic echo with relatively strict timing criteria (<3 cardiac cycles) for bubble appearance on the left side. However, failure to detect small defects was presumably evenly distributed between groups. Finally, the self-reporting of migraine could lead to potential misclassification. The results however are consistent with those of the MIST (Migraine Intervention with STARFlex Technology) trial that randomized 147 patients to percutaneous PFO closure or a sham procedure and showed no benefit in the primary end point—cessation of headache—at 6 months (29).
|
Valvular Heart Disease
|
|---|
Assessment of mitral regurgitation.
The proximal isovelocity surface area (PISA) has been widely used as a quantitative method for measuring the effective regurgitant orifice area (EROA) and mitral regurgitant stroke volume (MRSV) in patients with mitral regurgitation. This method assumes that flow approaches the orifice in an unconstrained manner and that there is a point orifice so that the streamlines of flow form a series hemispheric isovelocity surfaces as they approach the orifice. Measurement of the velocity and area of any surface in theory permits determination of the flow through the orifice (MRSV) and the EROA. Use of a single such surface to calculate the EROA also assumes that the orifice is constant throughout systole. Unfortunately, 1 or more of these assumptions may not be satisfied in individual patients. In a study to examine the effects of variations in mitral regurgitant flow rate (MRFR) and EROA on MRSV, Buck et al. (30) compared 4 different single-point and time-integrated PISA methods with similar MRI-derived data in 73 patients with mitral regurgitation (MR) of different etiologies. The MRSV was calculated from the apical 4-chamber view using 4 PISA methods: 1) the standard PISA approach in which the peak MRFR multiplied by the ratio of the velocity time integral to the maximum velocity equals the MRSV; 2) a simplified PISA formula multiplying single-point MRFR by an empirical factor of 3.25; 3) a time-integral PISA by the sum of serial PISA MRFR; and 4) a time-integral PISA from color M-mode recordings. Over all patients and mechanisms of MR, M-mode PISA correlated best with MRI with the smallest mean error (r = 0.88, SEE = 4.8, mean error –8.0 ± 6 ml) followed by serial PISA (r = 0.83, SEE = 5.9 ml, mean error –8.7 ± 7.4). For PISA-VTI, the values were r = 0.64, SEE = 7.4 ml, mean error = –13.3 ± 10.2 ml; and for simplified PISA methods they were r = 0.63, SEE = 7.4 ml, mean error –13.5 ± 10.3 ml. The investigators account for the systematic underestimation of MRSV by failure to account for the hemielliptic shape of many PISA jets (31). However, as they also note, there has been limited validation of the MRI method; therefore, although the data regarding the relative accuracy of these PISA methods are undoubtedly correct, the absolute differences may vary.
In another study, Matsumura et al. (32) sought to correct for the hemielliptical shape of the PISA as determined by real-time 3-dimensional echocardiography (RT3DE) in patients with functional MR using a computer program that included the horizontal width of the PISA along with 3 equally spaced vertical widths and radial dimensions. When compared with echo Doppler estimates of regurgitant volume, this 3-dimensional correction underestimated the regurgitant orifice area by only 26%, compared with 49% for the standard hemispheric measure. Future studies accounting for variability in both the spatial and temporal variability of the PISA will be required to ultimately determine the potential accuracy of this method (30).
Although PISA estimates the EROA based on the configuration and velocity of the flow stream approaching the orifice, a simpler alternative approach has been to use the diameter of the vena contracta of the flow exiting the regurgitant orifice as a measure of the EROA. Use of a single diameter again assumes a circular orifice, and the often complex shape of the regurgitant orifice, particularly in patients with functional MR, can invalidate this assumption. In a study of 57 patients with mild to severe MR of different etiologies, Kahlert et al. (33) compared the shape and size of the vena contracta by RT3DE with the 2-dimensional vena contact measured from apical 2- and 4-chamber diameters assuming circular and elliptical shapes. The RT3DE en face views of the vena contracta were successfully obtained in all patients. The vena contracta was noncircular in the majority of patients, with the ratio of the major to minor dimension averaging 1.18 ± 0.18 in the degenerative MR group, 1.58 ± 0.46 in the MVP group, and 2.87 ± 1.05 in the function MR group. Correlations of 2-dimensional vena contracta width in the apical 2- and 4-chamber views with the 3-dimensional area were modest, with the best correlation being obtained between the mean 2-dimensional vena contracta width and the 3-dimensional area (r = 0.90) with the smallest difference between the 3 different etiologies. There was a good correlation between vena contracta area by RT3DE and EROA by PISA using the hemispheric assumption (r = 0.93); however, there was significant underestimation of EROA by the PISA method (y = 0.62x + 0.04), particularly in patients with MVP and functional MR. Taking into account the hemielliptical shape of the PISA yielded a better correlation with VAC by RT3DE (r = 0.96) with less underestimation (y = 0.77x + 0.06). These data raise questions about current cutoff values for the severity of EROA by both PISA and vena contracta width. Although in the study by Buck et al. (30) (vide supra) the same group emphasized the importance of correcting for the temporal variation in PISA, such correction was not used in this study.
|
Bicuspid Aortic Valve
|
|---|
Bicuspid aortic valve is the most common congenital cardiac abnormality and is associated with a variety of complications, including aortic stenosis, insufficiency, and endocarditis. Bicuspid aortic valve is also associated with aortic dilation, which can progress to aneurysm formation and dissection. Because of the prevalence and potential complications of bicuspid aortic valve, an understanding of its natural history is important. Unfortunately, most data to date come from autopsy series or referral populations and hence are subject to selection bias. To examine the natural history in a community setting, Michelena et al. (34) studied 212 asymptomatic community residents (age 31 ± 20 years, 65% male, EF 50%) with no or mild aortic stenosis or regurgitation who were followed up for 15 ± 6 years. Three patients died of aortic valve disease (aortic stenosis [n = 1], aortic regurgitation [n = 1], and endocarditis [n = 1]). Overall survival was 97 ± 20% and 90 ± 3%, at 10 and 20 years, identical to the general population. Thirty-eight patients required aortic valve replacement (severe AS [n = 26], severe AR [n = 6], mixed valve disease [n = 2], moderate valve disease with severe aortic dilatation [n = 3], and subacute bacterial endocarditis [n = 1]) with a 20-year aortic valve surgery incidence of 24 ± 4%; 8 patients required coarctation repair (20-year rate 4 ± 1%); and 8 patients required surgery for ascending aortic dilatation or aneurysm (20-year rate 5 ± 2%). Ascending aorta dilatation (>40 mm) was noted in 15% at baseline and 39% at follow-up. No patient developed aortic dissection. During follow-up, congestive heart failure (CHF) occurred in 10 patients (7 ± 2%), 4 patients developed subacute bacterial endocarditis, and there were 5 strokes. New cardiovascular symptoms (dyspnea, syncope, or angina) occurred in 41 patients (26 ± 4% at 20 years). Comparison to the general population showed that although CHF rates were roughly similar, CHF occurred at an earlier age. However, the rate of aortic valve surgery was considerably higher and valve replacement occurred at an earlier age. Only age and degree of valve degeneration (thickening, calcification, and/or reduced mobility) at the initial examination were multivariate predictors of future cardiac events. Ascending aortic diameter 40 mm at baseline predicted subsequent aortic surgery independent of age and sex. Thus, despite relatively normal hemodynamics at baseline, the presence of valve degeneration indicates a poorer prognosis related primarily to progression of aortic stenosis (34).
In summary, reassessment and re-evaluation are essential in the evolution of diagnostic techniques. This is particularly important with techniques such as echocardiography in which: 1) methods that are validated and results reported in in vitro models and animal experiments are often difficult to reproduce in the more demanding clinical environment; 2) differences noted between selected groups in initial studies are less definitive when applied to the general population; and 3) assumptions inherent in quantitative techniques are often ignored when applied clinically. These difficulties are compounded when using methods that report quantitative data either directly or graphically from images (parametric imaging) but in which the underlying methods and calculations are invisible to the operator. However, with time, problems are identified and uses refined so that the important clinical and research question can be appropriately addressed.
|
References
|
|---|
1. Chung ES, Leon AR, Tavazzi L, et al. Results of the Predictors of Response to CRT (PROSPECT) trial Circulation 2008;117:2608-2616.[Abstract/Free Full Text]2. Marwick TH. Hype and hope in the use of echocardiography for selection for cardiac resynchronization therapy: the Tower of Babel revisited Circulation 2008;117:2573-2576.[Free Full Text] 3. Beshai JF, Grimm RA, Nagueh SF, et al. Cardiac-resynchronization therapy in heart failure with narrow QRS complexes N Engl J Med 2007;357:2461-2471.[CrossRef][Web of Science][Medline] 4. Lim P, Buakhamsri A, Popovic ZB, et al. Longitudinal strain delay index by speckle tracking imaging: a new marker of response to cardiac resynchronization therapy Circulation 2008;118:1130-1137.[Abstract/Free Full Text] 5. Ypenburg C, van Bommel RJ, Delgado V, et al. Optimal left ventricular lead position predicts reverse remodeling and survival after cardiac resynchronization therapy J Am Coll Cardiol 2008;52:1402-1409.[Abstract/Free Full Text] 6. Suffoletto MS, Dohi K, Cannesson M, Saba S, Gorcsan 3rd J. Novel speckle-tracking radial strain from routine black-and-white echocardiographic images to quantify dyssynchrony and predict response to cardiac resynchronization therapy Circulation 2006;113:960-968.[Abstract/Free Full Text] 7. Knebel F, Schattke S, Bondke H, et al. Evaluation of longitudinal and radial two-dimensional strain imaging versus Doppler tissue echocardiography in predicting long-term response to cardiac resynchronization therapy J Am Soc Echocardiogr 2007;20:335-341.[CrossRef][Web of Science][Medline] 8. Delgado V, Ypenburg C, van Bommel RJ, et al. Assessment of left ventricular dyssynchrony by speckle tracking strain imaging comparison between longitudinal, circumferential, and radial strain in cardiac resynchronization therapy J Am Coll Cardiol 2008;51:1944-1952.[Abstract/Free Full Text] 9. Notomi Y, Popovic ZB, Yamada H, et al. Ventricular untwisting: a temporal link between left ventricular relaxation and suction Am J Physiol Heart Circ Physiol 2008;294:H505-H513.[Abstract/Free Full Text] 10. Wang J, Khoury DS, Yue Y, Torre-Amione G, Nagueh SF. Left ventricular untwisting rate by speckle tracking echocardiography Circulation 2007;116:2580-2586.[Abstract/Free Full Text] 11. Tanaka H, Oishi Y, Mizuguchi Y, et al. Contribution of the pericardium to left ventricular torsion and regional myocardial function in patients with total absence of the left pericardium J Am Soc Echocardiogr 2008;21:268-274.[CrossRef][Web of Science][Medline] 12. Pettersen E, Helle-Valle T, Edvardsen T, et al. Contraction pattern of the systemic right ventricle shift from longitudinal to circumferential shortening and absent global ventricular torsion J Am Coll Cardiol 2007;49:2450-2456.[Abstract/Free Full Text] 13. Taber LA, Yang M, Podszus WW. Mechanics of ventricular torsion J Biomech 1996;29:745-752.[CrossRef][Web of Science][Medline] 14. Abraham TP, Pinheiro AC. Speckle-derived strain a better tool for quantification of stress echocardiography? J Am Coll Cardiol 2008;51:158-160.[Free Full Text] 15. Reant P, Labrousse L, Lafitte S, et al. Experimental validation of circumferential, longitudinal, and radial 2-dimensional strain during dobutamine stress echocardiography in ischemic conditions J Am Coll Cardiol 2008;51:149-157.[Abstract/Free Full Text] 16. Becker M, Lenzen A, Ocklenburg C, et al. Myocardial deformation imaging based on ultrasonic pixel tracking to identify reversible myocardial dysfunction J Am Coll Cardiol 2008;51:1473-1481.[Abstract/Free Full Text] 17. Remme EW, Lyseggen E, Helle-Valle T, et al. Mechanisms of preejection and postejection velocity spikes in left ventricular myocardium: interaction between wall deformation and valve events Circulation 2008;118:373-380.[Abstract/Free Full Text] 18. Mullens W, Borowski AG, Curtin RJ, Thomas JD, Tang WH. Tissue Doppler imaging in the estimation of intracardiac filling pressure in decompensated patients with advanced systolic heart failure Circulation 2009;119:62-70.[Abstract/Free Full Text] 19. Moller JE, Whalley GA, Dini FL, et al. Independent prognostic importance of a restrictive left ventricular filling pattern after myocardial infarction: an individual patient meta-analysis: Meta-Analysis Research Group in Echocardiography acute myocardial infarction Circulation 2008;117:2591-2598.[Abstract/Free Full Text] 20. Masutani S, Little WC, Hasegawa H, Cheng HJ, Cheng CP. Restrictive left ventricular filling pattern does not result from increased left atrial pressure alone Circulation 2008;117:1550-1554.[Abstract/Free Full Text] 21. Kusnetzky LL, Khalid A, Khumri TM, Moe TG, Jones PG, Main ML. Acute mortality in hospitalized patients undergoing echocardiography with and without an ultrasound contrast agent: results in 18,671 consecutive studies J Am Coll Cardiol 2008;51:1704-1706.[Abstract/Free Full Text] 22. Wei K, Mulvagh SL, Carson L, et al. The safety of Definity and Optison for ultrasound image enhancement: a retrospective analysis of 78,383 administered contrast doses J Am Soc Echocardiogr 2008;21:1202-1206.[CrossRef][Web of Science][Medline] 23. Behm CZ, Kaufmann BA, Carr C, et al. Molecular imaging of endothelial vascular cell adhesion molecule-1 expression and inflammatory cell recruitment during vasculogenesis and ischemia-mediated arteriogenesis Circulation 2008;117:2902-2911.[Abstract/Free Full Text] 24. Villanueva FS, Lu E, Bowry S, et al. Myocardial ischemic memory imaging with molecular echocardiography Circulation 2007;115:345-352.[Abstract/Free Full Text] 25. Coli S, Magnoni M, Sangiorgi G, et al. Contrast-enhanced ultrasound imaging of intraplaque neovascularization in carotid arteries: correlation with histology and plaque echogenicity J Am Coll Cardiol 2008;52:223-230.[Abstract/Free Full Text] 26. Shah F, Balan P, Weinberg M, et al. Contrast-enhanced ultrasound imaging of atherosclerotic carotid plaque neovascularization: a new surrogate marker of atherosclerosis? Vasc Med 2007;12:291-297.[Abstract/Free Full Text] 27. Galiuto L, Garramone B, Scara A, et al. The extent of microvascular damage during myocardial contrast echocardiography is superior to other known indexes of post-infarct reperfusion in predicting left ventricular remodeling: results of the multicenter AMICI study J Am Coll Cardiol 2008;51:552-559.[Abstract/Free Full Text] 28. Rundek T, Elkind MS, Di Tullio MR, et al. Patent foramen ovale and migraine: a cross-sectional study from the Northern Manhattan Study (NOMAS) Circulation 2008;118:1419-1424.[Abstract/Free Full Text] 29. Dowson A, Mullen MJ, Peatfield R, et al. Migraine Intervention With STARFlex Technology (MIST) Trial: a prospective, multicenter, double-blind, sham-controlled trial to evaluate the effectiveness of patent foramen ovale closure with STARFlex septal repair implant to resolve refractory migraine headache Circulation 2008;117:1397-1404.[Abstract/Free Full Text] 30. Buck T, Plicht B, Kahlert P, Schenk IM, Hunold P, Erbel R. Effect of dynamic flow rate and orifice area on mitral regurgitant stroke volume quantification using the proximal isovelocity surface area method J Am Coll Cardiol 2008;52:767-778.[Abstract/Free Full Text] 31. Yosefy C, Levine RA, Solis J, Vaturi M, Handschumacher, MD, Hung J. Proximal flow convergence region as assessed by real-time 3-dimensional echocardiography: challenging the hemispheric assumption J Am Soc Echocardiogr 2007;20:389-396.[CrossRef][Web of Science][Medline] 32. Matsumura Y, Saracino G, Sugioka K, et al. Determination of regurgitant orifice area with the use of a new three-dimensional flow convergence geometric assumption in functional mitral regurgitation J Am Soc Echocardiogr 2008;21:1251-1256.[CrossRef][Web of Science][Medline] 33. Kahlert P, Plicht B, Schenk IM, Janosi RA, Erbel R, Buck T. Direct assessment of size and shape of noncircular vena contracta area in functional versus organic mitral regurgitation using real-time three-dimensional echocardiography J Am Soc Echocardiogr 2008;2:912-921. 34. Michelena HI, Desjardins VA, Avierinos JF, et al. Natural history of asymptomatic patients with normally functioning or minimally dysfunctional bicuspid aortic valve in the community Circulation 2008;117:2776-2784.[Abstract/Free Full Text]
Related Article
-
Inside This Issue
J. Am. Coll. Cardiol. 2009 53: A28.
[Full Text]
[PDF]
This article has been cited by other articles:
|
|
|
|
 
L. H. Eraso, M. P. Reilly, C. Sehgal, and E. R. Mohler
Emerging diagnostic and therapeutic molecular imaging applications in vascular disease
Vascular Medicine,
April 1, 2011;
16(2):
145 - 156.
[Abstract]
[PDF]
|
|
|
|
|
|
|
Y. Mizuguchi, Y. Oishi, H. Miyoshi, A. Iuchi, N. Nagase, N. Ara, and T. Oki
Possible mechanisms of left ventricular torsion evaluated by cardioreparative effects of telmisartan in patients with hypertension
Eur Heart J Cardiovasc Imaging,
September 1, 2010;
11(8):
690 - 697.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
Top
CRT
Speckle Tracking for CRT