Magnetic Resonance Imaging for Congenital Heart Disease

Pulse Sequences

ECG-gated spin echo and gradient echo sequences are used for cardiac evaluation. Spin echo, or “black blood,” imaging gives excellent contrast resolution between the endocardium or vessel wall and the blood-filled lumen ( Fig. 67.1 ). Alternatively, on gradient echo (GRE) images, or “bright blood” images, the blood pool has high signal intensity. The GRE sequence balanced SSFP (b-SSFP) is the current mainstay of cardiac imaging. Balanced SSFP sequences are faster than (GRE) sequences and yield better contrast between myocardium and blood. By obtaining data during set phases of the cardiac cycle, b-SSFP cine images can be used for analysis of myocardial or valvular motion throughout the cardiac cycle. Although b-SSFP is ideally performed with the patient holding his or her breath, respiratory motion artifacts are minimized in sedated or uncooperative children by increasing the number of excitations. This technique is the preferred sequence for evaluating cardiac motion and quantifying cardiac function ( Fig. 67.2 ).

Functional Evaluation

MRI is the reference standard for noninvasive cardiac functional assessment and is superior to echocardiography in assessing right ventricular function, particularly in patients with congenital heart disease. This is an important consideration in evaluation of these children as they generally need serial and reproducible measurements of ventricular function to determine optimal timing of potential interventions. A stack of short-axis cine images through the ventricles is obtained to measure ventricular volumes. Using postprocessing software, the operator defines the endocardial border on each slice, and the software then multiplies the resultant cross-sectional area by the slice thickness to obtain the ventricular volume. There is no consensus on inclusion or exclusion of the papillary muscles in the left ventricular volume or trabeculation in the right ventricular volume, so either method should be used with knowledge that the volumes can be affected by differences in postprocessing technique and similar methodology should be used when comparing studies. End-diastolic and end-systolic phases are identified for calculation of ejection fraction and stroke volume. The additional application of pericardial contours allows determination of myocardial mass.

Because children vary in size, cardiac MRI measurements are reported as indexed values. These are obtained by dividing each measurement by the patient's body surface area. This allows more standardized comparisons as patients grow and across patient populations. There are reported normal values for left and right ventricular volumes in children, and these can be used for general reference but are based on small sample sizes.

The right and left ventricular stroke volumes are the same in the structurally normal heart. Differences in stroke volumes are caused by shunts and/or valvular regurgitation. Using MRI, both shunt volumes and regurgitant fractions can be calculated from stroke volumes or flow data if there is an isolated single shunt lesion or regurgitation involving a single valve. If there are multiple abnormalities, more detailed evaluation is needed. Shunt fraction, or the pulmonary to systemic flow ratio (Qp:Qs), is determined by dividing the larger stroke volume by the smaller. The valvular regurgitant fraction is determined by dividing the regurgitant volume, which is the difference of ventricular stroke volumes, by the larger ventricle stroke volume ( Fig. 67.3 ). The regurgitant volume may result from atrioventricular valve or semilunar valve insufficiency, or a combination of both.

where Ventricular stroke volume ₁ is the larger of the two ventricular stroke volumes.

Phase-contrast imaging uses a GRE sequence to measure phase shift and produces both magnitude and directional information ( Fig. 67.4 ). By setting an acquisition plane perpendicular to a vessel or valve of interest, phase-contrast imaging allows for accurate quantification of blood flow velocity. The pressure gradient across a stenosis can then be calculated from the blood flow velocity using the modified Bernoulli equation:

where Δ P is the peak instantaneous pressure gradient across the stenosis and v is the velocity across the stenosis measured by flow analysis. Using postprocessing software, the velocity within each pixel of the targeted vessel can be multiplied by the vessel area and then summed across the cross section of the vessel across the cardiac cycle to generate the flow per beat within the vessel. Given the limitless scan plans of cardiac MRI, this method has proved superior to echocardiography for flow determinations and is used extensively in congenital heart disease to evaluate stenoses, shunt fractions, and collateral vessel burden. More recently, 3D phase-contrast imaging (four-dimensional [4D] flow) has allowed acquisition of a volumetric 3D flow dataset over the cardiac cycle, which can be postprocessed in any plane to yield velocity and flow data throughout the acquired volume.