Feasibility of tissue magnetic resonance imaging
A pilot study in comparison with tissue Doppler imaging and invasive measurement
Bernard P. Paelinck, MD*,*,
Albert de Roos, MD, PhD ,
Jeroen J. Bax, MD, PhD ,
Johan M. Bosmans, MD, PhD*,
Rob J. van Der Geest, MSc,
Dominique Dhondt, MSc*,
Paul M. Parizel, MD, PhD ,
Christiaan J. Vrints, MD, PhD* and
Hildo J. Lamb, MD, MSc, PhD,*
* Department of Cardiology, University Hospital Antwerp, Antwerp, Belgium
Department of Radiology, University Hospital Antwerp, Antwerp, Belgium
Department of Radiology, Leiden University Medical Center, Leiden, the Netherlands
Department of Cardiology, Leiden University Medical Center, Leiden, the Netherlands
View larger version (20K):
[in a new window]
|
Figure 1 Diagram of mitral filling pattern, mitral annulus velocities, and corresponding four-chamber view during early diastole. (Left side) Left ventricular (LV) diastole is characterized by swift myocardial relaxation and elastic recoil. As a result, normal transmitral flow is characterized by a prominent and rapid early (E) filling wave (arrow) due to passive suction, and by a diminutive late (A) atrial filling wave due to atrial contraction. During diastole, the LV expands longitudinally resulting in a descent of the mitral valve annulus (open arrow) as opposed to the relatively fixed apex. As the velocity of the earliest diastolic motion of the mitral valve annulus (Ea) relates to the rate of myocardial relaxation, Ea is prominent in normal hearts. (Right side) When diastolic dysfunction occurs, myocardial relaxation is impaired; LV pressure falls slowly, reducing early transmitral driving pressure. As a result, E is decreased (arrow), and, because of increased atrial preload, A is increased. Due to decreased myocardial relaxation velocity, Ea is reduced (open arrow). As disease progresses, LV compliance also becomes impaired, and LV filling becomes dependent on increased left atrial (LA) or filling pressure (dashed arrow). This results in an increased E. As the underlying impaired relaxation is masked and because the pattern resembles the normal filling pattern, it is called pseudonormalized pattern. Combining E, which is dependent on both filling pressure and myocardial relaxation, with Ea, which is mainly dependent on myocardial relaxation, allows differentiation of a normal from a pseudonormal signal and better evaluation of filling pressures (see equation). E = peak mitral velocity in early diastole; Ea = early diastolic tissue velocity.
|
|
View larger version (66K):
[in a new window]
|
Figure 2 (A) Planscan for acquisition of transmitral flow and tissue velocities. To assess transmitral flow, a phase-contrast magnetic resonance (MR) sequence was used, with a velocity encoding of 130 cm/s and the center of the slice positioned perpendicular to mitral inflow, at early diastole (upper panel). To assess tissue MR velocities, phase-contrast MR was repeated with a velocity encoding of 30 cm/s, and the image slice positioned at two-thirds of the long axis, planned on early diastolic two- and four-chamber images, perpendicular to the interventricular septum (lower panel). (B) Velocity-encoded images of transmitral flow (upper panel) and tissue velocities (lower panel). The tissue velocities are measured from a circular region-of-interest of 20 pixels in the posteroseptal region. (C) Corresponding velocity versus time curves. From these curves, peak mitral velocity in early diastole (E) = 103 cm/s, early diastolic posteroseptal tissue velocity Ea = 8.9 cm/s and E/Ea = 11.6 were derived. Ea = early diastolic tissue velocity; Venc = velocity encoding.
|
|
View larger version (125K):
[in a new window]
|
Figure 3 A standardized circular region of interest (ROI) of 20 pixels was placed at different locations around the circumference of the heart. A reference line (II) was traced through the center of the left ventricle (LV), and the intersection of the posterior LV with the right ventricle (RV). Based on this reference line, the LV myocardium was divided into three equal triangular regions, corresponding with the coronary artery territories; ROIs were placed in the center of the myocardium at the following locations: posteroseptal (1), midseptal (2), anteroseptal (3), lateral (4), and inferior (5). The posteroseptal ROI was positioned at 30 degrees posterior to line I.
|
|
View larger version (16K):
[in a new window]
|
Figure 4 (A) Correlation of magnetic resonance and Doppler-measured E/A and E/Ea. (B) Bland-Altman plot of the difference between magnetic resonance (MR) and Doppler-measured E/A and between posteroseptal MR and Doppler-measured E/Ea. Horizontal lines show the mean difference between two methods ± 2 SDs. A = peak mitral velocity at atrial contraction; E = peak mitral velocity in early diastole; Ea = early diastolic tissue velocity.
|
|
View larger version (15K):
[in a new window]
|
Figure 5 (A) Correlation between magnetic resonance (MR)-measured E/A and invasive mean pulmonary capillary wedge pressure (PCWP). Note absence of correlation (p = 0.99). (B) Correlation between posteroseptal MR-measured E/Ea and invasive PCWP (R2 = 0.65, p < 0.0001). (C) Bland-Altman plot of the difference between MR-estimated and catheter-measured PCWP. Horizontal lines show the mean difference between two methods ± 2 SDs. A = peak mitral velocity at atrial contraction; E = peak mitral velocity in early diastole; Ea = early diastolic tissue velocity.
|
|
View larger version (17K):
[in a new window]
|
Figure 6 Bland-Altman plots illustrating good intraobserver and interobserver variability for posteroseptal magnetic resonance (MR)-measured E/Ea. Horizontal lines show the mean difference between two analyses ± 2 SDs. E = peak mitral velocity in early diastole; Ea = early diastolic tissue velocity.
|
|
|
