A new method for cardiac computed tomography regional function assessment: stretch quantifier for endocardial engraved zones (SQUEEZ)

Abstract

Background: Quantitative assessment of regional myocardial function has important diagnostic implications in cardiac disease. Recent advances in CT imaging technology have allowed fine anatomic structures, such as endocardial trabeculae, to be resolved and potentially used as fiducial markers for tracking local wall deformations. We developed a method to detect and track such features on the endocardium to extract a metric that reflects local myocardial contraction.

Methods and results: First-pass CT images and contrast-enhanced cardiovascular magnetic resonance images were acquired in 8 infarcted and 3 healthy pigs. We tracked the left ventricle wall motion by segmenting the blood from myocardium and calculating trajectories of the endocardial features seen on the blood cast. The relative motions of these surface features were used to represent the local contraction of the endocardial surface with a metric we call stretch quantifier of endocardial engraved zones (SQUEEZ). The average SQUEEZ value and the rate of change in SQUEEZ were calculated for both infarcted and healthy myocardial regions. SQUEEZ showed a significant difference between infarct and remote regions (P<0.0001). No significant difference was observed between normal myocardium (noninfarcted hearts) and remote regions (P=0.8).

Conclusions: We present a new quantitative method for measuring regional cardiac function from high-resolution volumetric CT images, which can be acquired during angiography and myocardial perfusion scans. Quantified measures of regional cardiac mechanics in normal and abnormally contracting regions in infarcted hearts were shown to correspond well with noninfarcted and infarcted regions as detected by delayed enhancement cardiovascular magnetic resonance images.

Figure 1.

Steps of the proposed method. (A) cropped axial CT image of the LV. (B) the blood pool segmented from the volume by thresholding. (C) Endocardial surface extracted from the segmented images (inferior wall facing viewer). (D) shape index values calculated to encode the features engraved by the trabecular structures on the endocardial surface. Coherent point drift algorithm is used to find the correspondence between the endocardial features at end diastole (ED) (left), used as template, and other systolic phases (right). (E) SQUEEZ is calculated for each triangle on the ED endocardial surface mesh by tracking the corresponding triangle at different cardiac phases. A(0) is the area of the triangle at ED, and A(t) is its area at cardiac phase t. (F) SQUEEZ maps calculated for every triangle on the endocardial surface at five cardiac phases from end diastole to end systole.

Figure 2.

Local shape encoding using shape index. Left: Shape index values for the point at the center of the simple surfaces. From top : spherical cap, ridge, saddle point, valley, and spherical cup. Spherical cap: k1 = k2 >0 thus SI = +1; ridge: k1>0 ,k2=0 and SI=+0.5; saddle point k1=−k2≠0, SI=0; valley: k1=0, k2<0 thus SI = −0.5; spherical cup: k1=k2<0 results in SI=−1; other SI values are caused by smooth deformation of these surfaces. Right: an example of shape index calculated for an ED endocardial surface.

Figure 3.

Global left ventricle function measures for healthy (n=3) and infarcted (n=8) pigs. From left to right: end diastolic volume (EDV): 87.8±17.7 (healthy) and 92.5±15.6 (infarcted); end systolic volume (ESV): 40.1±7.3 (healthy) and 50.4±6.6 (infarcted); stroke volume (SV= EDV-ESV): 47.6±10.5 (healthy) and 42.0±11.0 (infarcted) ; all in milliliters, and ejection fraction percentage (E_f= SV/EDV): 54.1±1.3 (healthy) and 44.9±5.6 (infarcted). The bars and whiskers indicate the mean and ± standard deviation of the quantities, respectively. (* indicates p<0.05)

Fig.4.

Bull’s-eye plots of the SQUEEZ values for 3 typical infarcted (A), and 3 healthy pigs (B) from end diastole (left) to end systole (right) at 10% R-R steps. Infarcted subjects show abnormal stretching of the endocardium in LAD territory (anterior and anteroseptal segments) which is consistent with the infarction model (LAD occlusion after the second diagonal) used in this study and the locations observed with contrast enhanced MR.

Fig.5.

(A) Left: A short axis phase-sensitive inverted recovery (PSIR) MR image of an animal with an anterior/anteroseptal heterogeneously infarcted region. Infarcted region has characteristic high signal intensity. Right: End systolic SQUEEZ bull’s-eye plot of the same animal. The short axis image in the left approximately corresponds to the SQUEEZ values along the dashed arc. The infarcted sub-regions in the MR image correspond to the regions detected in the SQUEEZ plot, depicted by the black arrows. Showing from left to right, a section with some loss of function, a section with complete loss of function that shows wall expansion, a small section with some contractility and a fourth sub-region with loss of function. (B) Time plots of the average SQUEEZ values for healthy, MI, and non-MI regions in systole for 3 healthy (top row) and 7 infarcted pigs (middle and bottom rows). The regions were chosen to be roughly the size of segments in the AHA 17-segment model. All infarcted pigs showed significant difference in SQUEEZ for MI and non-MI regions (p<0.0001). ( ** denotes the dose-modulated dataset) (C) Average SQUEEZ rate values calculated by averaging over the slopes of lines fitted to the curves in (B). SQUEEZ rate is significantly different between MI and non-MI regions in infarcted hearts (p<0.0001). There was no significant difference between non-MI regions in the infarcted hearts and the same regions chosen in the healthy hearts.

Figure 6-

Sample comparison between high dose and low dose scans: Bland-Altman plot of the SQUEEZ values calculated from high-dose retrospective and low-dose prospective scans shows low bias (mean= 0.01) and standard deviation = 0.07. Dashed lines denote 95% confidence interval CI= [−0.12 0.15]. (N=2250)