Physical Address
304 North Cardinal St.
Dorchester Center, MA 02124
Assessment of Left Ventricular Function, Infarction, Perfusion, and Viability
Assessment of Left Ventricular Volumes And Systolic Function
The excellent spatial resolution of contrast-enhanced cardiac CT (CCT) permits the delineation of accurate endocardial and epicardial borders to a degree exceeding that of cardiac MRI. The avoidance of epicardial chemical shift artifact and the clear distinction between blood pool and myocardium (due to less blurring from partial volume effects and elimination of flow artifacts on cine gradient echo and steady-state free precession images in regions of increased trabeculation or slow flow) facilitates the straightforward quantification of left ventricular (LV) volumes and function.
By reconstructing retrospectively gated images, CCT is able to depict the regional and global systolic and diastolic phases of the heart and also to:
- □
Run as a cine-loop that allows for regional and overall wall motion assessment
- □
Calculate, using biplane Simpson’s method applied to four- and two-chamber views, or the method of disks applied to a short-axis stack:
-
Diastolic volume
-
Systolic volume
-
Stroke volume
-
Ejection fraction
-
- □
Calculate and display:
-
Diastolic wall thickness
-
Systolic wall thickness
-
Systolic wall thickening
-
Systolic regional motion
-
Use of automatic edge detection and chamber segmentation algorithms and Simpson’s method enables determination of LV end-diastolic volume from a selected diastolic reconstruction and LV end-systolic volume from a selected systolic reconstruction.
Selection of End-systole and End-diastole
End-systole
Although end-systole often is 35% or 40% of the R-R interval, that is not always the case, especially when the heart rate is elevated. Changing the reconstruction by even 1% can influence the determination of end-systolic volume. To define end-systole optimally (given contemporary imaging temporal resolution of the acquisition, which determines the number of required phases needed), more rather than fewer phases should be used, because the volumetric assessment depends on the number of cardiac cycles. Misidentification of end-systole increases measured end-systolic volume, by up to 20%. At least 20 reconstructed phases should be used.
End-diastole
Use of 0% R-R interval should achieve an accurate end-diastole representation.
Use of mid-diastole as for coronary CT imaging is not an accurate determination of end-diastole, except possibly in atrial fibrillation, because the atrial contribution to ventricular filling would not yet have occurred.
Segmentation Algorithms
Many algorithms have been developed that assist with chamber or infarct segment delineation, based on attenuation differences. Chamber identification/segmentation algorithms are based on the expected shape of the cavities of the cardiac chambers and great vessels and reduce the time to yield quantitation. At the same time, however, these algorithms increase variability when compared with manually traced short-axis slices. Often some manual adjustment still is required to obtain accurate chamber delineation. Typically, semi-automated quantification of LV volumes and function takes about 5 minutes with contemporary post-processing software. Post-processing time can also be reduced by segmenting only end-diastolic and end-systolic images instead of data from the entire cardiac cycle, as is usually done in cardiac MR (CMR). Semiautomated segmentation also may be improved by reconstructing the functional data set at an increased slice thickness (1.2–3 mm). This may reduce the amount of manual adjustment that may subsequently be needed.
Segmentation of infarct territory using a delayed enhancement technique is based on an arbitrary number of standard deviations above sampling of non-infarct territory.
As with all imaging modalities, navigating the complexities of the basal anatomy of the ventricles, especially the shape of the atrioventricular valve and of the outflow tract, is a challenge and generally requires manual editing.
You're Reading a Preview
Become a Clinical Tree membership for Full access and enjoy Unlimited articles
If you are a member. Log in here
