Continuous Mechanical Wave Imaging: Towards automated cardiac stiffness imaging
Résumé
Background, Motivation and Objective
Myocardial fibrosis is known as a physio-pathologic substrate of several cardiovascular diseases.
Knowing that fibrosis causes increased myocardial stiffness, natural mechanical wave (MW) imaging has
shown potential for detection of fibrotic tissue. Conventionally the MW speed is estimated using the slope
of a 2D spatiotemporal map obtained by selecting an anatomic m-mode line along the myocardium. This
procedure has to be done for each MW within a cardiac cycle which can be cumbersome and operator
dependent. Here, we propose a continuous MW imaging approach that allows estimating and mapping
the velocity of several MW (AVC, MVC, and AK) automatically.
Statement of Contribution/Methods
A 2D high frame rate setup was implemented on GE Vivid E95 with the M5Sc-D phased array, by
transmitting 6 steered planes waves (820 images/s). MWs were detected using the clutter filter wave
imaging technique. 2D time propagation maps were estimated by detecting the maxima of the waves front.
Assuming the MW follows the shorter path between two spatial points, the 2D local MW velocity was
estimated by computing the 2D gradient of the time propagation map. This procedure was performed in a
sliding temporal window of 80 ms, along the full ultrasound sequence. The MW speed was then mapped
on the B-mode images (A). N=13 healthy volunteers and N=18 patients with aortic stenosis were
investigated.
Results/Discussion
An example is shown in B (blue curve) for a specific region of interest. Interestingly, several flat regions
(red curve) emerged and matched (for the most part) at mechanical wave events. The size of the flat
regions is linked to the MW propagation time. As expected the estimated AVC, MVC, and AK velocities
were higher than the AK MW velocity (C). Moreover, the method distinguishes the healthy volunteer
group from the patient with all the mechanical waves produced by the heart. In this work, we presented a
method able to estimate simultaneously, locally and automatically the velocity of several MW produced
by the heart
Domaines
Sciences de l'ingénieur [physics]Origine | Fichiers produits par l'(les) auteur(s) |
---|