Cardiac complications are present in most muscular dystrophies. Although the range of injurious mechanisms is extremely varied, including defined modulating factors (eg, putative genes, signaling, and maintenance), age (or disease progression) at diagnosis and initial treatment, and treatment dosing, the need to characterize disease features with high accuracy is paramount. Because MRI can provide precise chamber volumes from three-dimensional images, and a wide range of other functional metrics, it is considered the measurement gold standard. Paterson and colleagues describe methods other than MRI (eg, CT, positron emission tomography, single photon emission CT). Because dystrophinopathies (eg, Duchenne muscular dystrophy, Becker muscular dystrophy, X-linked dilated cardiomyopathy) represent most of the published work incorporating MRI strain investigation, this literature is summarized to emphasize the current and future role of MRI in disease management.

MRI images focused on anatomy are typically acquired during the diastolic portion of the cardiac cycle to eliminate cardiac motion, with breath-holding generally used to eliminate respiratory artifacts. Multiple averaged acquisitions are another approach to ameliorating respiratory artifacts in patients who cannot perform breath-holds, as is often the case with advanced stages of muscular dystrophy. Specific modification of the sequence parameters produces longitudinal (T1) or transverse relaxation (T2) weighting, which is helpful for characterizing the myocardial tissues. Black blood sequences have been used traditionally for anatomic analysis and generally use electrocardiogram-gated spin echo techniques with successive nonselective and slice-selective inversion pulses, allowing suppression of the blood pool signal with restoration of myocardial signal before image acquisition.

The current workhorse for functional cardiac MRI is the balanced steady-state free precession (SSFP) gradient echo technique, which mainly replaced spoiled gradient echo imaging. SSFP sequences have bright-blood characteristics from T2/T1 effects, and image acquisition is segmented and spread over multiple heartbeats with the resulting cine allowing assessment of the entire cardiac cycle. Myocardial motion is easily depicted with cine sequences typically yielding a temporal resolution in the range of 50 ms (ie, 20 frames per cardiac cycle at a heart rate of 60 beats per minute), which provides a method for subjective assessment of myocardial contractility and wall motion abnormalities. Higher temporal resolutions can be used for valve information. Ventricular contours are traced on individual end systolic and end diastolic images to calculate ventricular volumes using the modified Simpson’s method ( Fig. 1 ). These volumes are used to derive the following usual parameters of ventricular function:

• 
  

  ESV: end systolic volume

  
  
• 
  

  EDV: end diastolic volume

  
  
• 
  

  SV: stroke volume = EDV – ESV

  
  
• 
  

  EF: ejection fraction = (SV/EDV) × 100

  
  
• 
  

  Cardiac output; SV × heart rate

End diastolic ( A ) and end systolic ( B ) short-axis SSFP images with endocardial contour tracings depicting right ( blue ) and left ( red ) ventricles.

Phase contrast techniques use gradient echo techniques to derive quantitative velocity and flow results from the blood pool within the vessel of interest. Various permutations of this technique are helpful in identifying flow disturbances, such as with stenosis, valvular pathology, and intracardiac shunts.

Administration of gadolinium-based contrast agents has efficacy in identifying and characterizing myocardial pathology. Typical gadolinium chelates accumulate within the extracellular space where they shorten T1 and T2 relaxation, with delayed clearance in areas of fibrosis or impaired vascularity. This detail forms the basis of late (delayed) gadolinium enhancement (LGE) techniques now widely used for detecting myocardial infarction and various cardiomyopathies. An example of LGE in a patient with Duchenne muscular dystrophy is shown in Fig. 2 .

Postcontrast phase-sensitive inversion-recovery (PSIR) true fast imaging with steady-state free precession (FISP) short-axis images showing subepicardial, transmural, and papillary muscle delayed myocardial enhancement in a 17-year-old patient with Duchenne muscular dystrophy ( white arrows ).

Various methods of strain imaging provide an alternate way to quantify myocardial tissue and wall deformation. Most commonly, several slices are sampled in the short-axis plane. The most reported metric derived from these methods is circumferential strain (ε _cc ), which is the shortening/wringing motion during systole, whereas other strain parameters, such as radial and longitudinal deformation, are also used. Furthermore, some studies focus on diastolic strain metrics. MRI parameters are principally in units of magnitude, although they can be reformatted to look at coherent changes in time, which is discussed in more detail.

Since its development approximately 20 years ago, strain assessment has been applied to a range of conditions and questions, such as patients with thalassemia ; a healthy cohort to determine relationship between cardiac parameters (eg, volumes) and strain metrics ; and patients after heart transplant. Widespread application has not been realized, however, probably because of the extra time required for examination (making it difficult in ill patients) and the limited availability of free or low-cost postprocessing methods.

Three main analysis methods are described in the literature: harmonic phase tagging (HARP), which is the most common; fast cine displacement-encoded (DENSE); and cine sequence-based feature tracking (FT ). Processed data then evaluate directional changes over time in cross-hatched “tags,” whereas in DENSE, the information is contained in opposing field gradients. Fig. 3 provides an example of a tagged image. Although HARP/DENSE methods are more similar than different, DENSE is often used for applications focused on greater in-plane resolution, at the expense of signal-to-noise per unit time. FT uses standard cine magnitude images plus segmentation to generate a strain map of the heart walls, with the caveat that this approach excludes a within-tissue component.

Example tags and strain metrics in a pediatric patient using fast cine displacement-encoded (DENSE).

(Software Courtesy of Auckland MRI Research Group; with permission.)

Duchenne muscular dystrophy, the most frequent (6/100,000 live births ) and severe of the dystrophinopathies, has been the focus of most articles using cardiac MRI, and strain specifically. Hermans and colleagues provide an excellent survey of broadly focused muscular dystrophy genetics, with attention to cardiac findings. Dystrophin is a large protein that composes part of the normal muscle membrane. In Duchenne muscular dystrophy, the entire reading frame is absent, resulting in complete protein absence, whereas in Becker muscular dystrophy (affecting 2.4/100,000 live births ) the reading frame remains, resulting in abnormally low or variable protein levels. In contrast, X-linked dilated cardiomyopathy does not result in skeletal muscle symptoms, although a rapidly progressive fatal cardiomyopathy occurs in the second decade of life.

The functional consequence of these membrane gaps is that muscle becomes highly vulnerable to injury during normal function. Because of the dramatic injury progression over time in these diseases, research investigation has focused on the early detection of subclinical vulnerabilities in the heart, with the hope that treatment can be initiated to delay or forestall life-threatening consequences.