bromodomain One of the main parameters derived from

One of the main parameters derived from LOKI analysis is its maximal peak-to-peak amplitude ΔX (i.e., the total amplitude of the longitudinal motion of the intima and media layers along the direction parallel to the blood flow), which is significantly reduced in patients with presumably stiffer bromodomain (Svedlund et al. 2011; Zahnd et al. 2011a, 2012). Yet, the influence of the location of the assessed ROI on the resulting motion amplitude ΔX is still unclear. Indeed, it has previously been reported by our team that different resulting trajectories could be observed when tracking points in different regions (i.e., on the right, center and left side of the image) (Zahnd et al. 2011a).
The aim of the pilot study described here was to characterize such local variability in LOKI amplitude ΔX in longitudinal B-mode US cine loops of in vivo healthy CCAs. For each cine loop analyzed, the motion of the far wall at different locations within the intima–media complex is evaluated on a collection of points by means of robust speckle tracking, using a previously validated framework (Zahnd et al. 2013). Our rationale is as follows: mechanical LOKI-inducing forces are likely to be progressively dissipated along the propagation in the artery, reflecting the elastic damping role of the circulatory system. Therefore, it is expected that ΔX is systematically lower in ROIs located on the “distal” side of the image (i.e., toward the head) compared with ROIs located on the “proximal” side (i.e., toward the heart). To the best of our knowledge, this phenomenon has not previously been observed in vivo.


The average LOKI amplitude ΔX of all individual tracked points was 722 ± 284 μm. As the number of tracked points is subject-dependent, we also define as the mean value of all ΔX measurements for a given participant. The average amplitude for all participants was 717 ± 260 μm. For the majority of the patients analyzed (i.e., 33/35), we made the following observation: LOKI amplitude ΔX is systematically larger for points located on the proximal side of the image and smaller for points located on the distal side of the image, as depicted in Figure 2.
Examples of the attenuation coefficient Z estimated in different participants are displayed in Figure 3. Calculating the average attenuation coefficient Z of all 35 participants, we confirmed the presence of a progressive attenuation of LOKI amplitude toward the distal direction, with an average Z value of −17 ± 16 μm/mm (range: −91 to 17 μm/mm), as illustrated in Figure 4. This attenuation can also be quantified as follows: within a given CCA segment, the motion between two points separated by 1 mm undergoes a decrease corresponding to −2.5 ± 2.0% of the average LOKI amplitude of the vessel. Arteries with a larger LOKI amplitude were not found to also undergo a stronger attenuation, as no significant correlation could be observed between the mean LOKI amplitude per subject, , and the corresponding attenuation coefficient Z (R = 0.22).
The average dispersion error with respect to the linear regression model (i.e., mean absolute vertical distance between the estimated points and the regression line) as a function of the number of tracked points is outlined in Table 1. The overall average absolute error of fitting between all the measurement points and the linear regression model was 34 ± 36 μm. By definition, the dispersion error generated by linear regression is null in cine loops with exactly two tracked points. For lymphatic system reason, we also calculated the average absolute error of fitting in cine loops with at least three measurement points, which was 39 ± 36 μm. Such dispersion remains small in comparison to the average LOKI amplitude (i.e., 717 ± 260 μm), as depicted in Figure 3. The dispersion was also nearly two times smaller than the tracking accuracy of the present framework in the longitudinal direction, previously validated in Zahnd et al. (2013). Indeed, with respect to the reference that was generated from the manual tracings of three experienced observers, the average absolute error of the present framework for ΔX quantification was 74 ± 68 μm, whereas the inter- and intra-observer variability was 79 ± 103 and 62 ± 103 μm, respectively.