Cardiac Multifrequency Time-harmonic Elastography: In Silico Validation and In Vivo Application.

Abstract

Purpose: To introduce and validate optimized multifrequency cardiac time-harmonic elastography (THE) with portable drivers and multiple-heartbeat acquisitions for detecting abnormal diastolic stiffness using in silico testing and investigating participants with confirmed cardiac wild-type transthyretin amyloidosis (ATTRwt), a model disease for myocardial stiffness, compared to healthy controls.

Materials and methods: Cardiac THE was performed using continuous vibrations of 60, 70, and 80 Hz induced by a portable driver. Shear wave speed (SWS) maps, as a proxy for myocardial stiffness, were acquired over 4 seconds in a parasternal long axis view. Cardiac ultrasound simulations with CT-data-based anatomies were used for in silico validation and parameter optimization. Diagnostic performance was assessed using the area under the ROC curve (AUC) with 95% confidence interval (95%-CI).

Results: Healthy controls (N=10; median age: 32 years [range: 27-56]; five men) and participants with ATTRwt (N=11; median age: 83 years [range: 64-86]; ten men) were investigated, resulting in accurate amyloidosis detection (AUC: 1.00; 95%-CI: 0.94-1.00). In-silico-estimated SWS agreed with ground-truth values (r²=0.94; median error: 5.6% [range: -26.2-19.7]) across various stiffnesses [range: 1.3-4.0 m s^-1] and thicknesses [range: 10-20 mm]. A trend towards underestimation was only observed in the unlikely scenario of thin and very stiff septa.

Conclusion: For the first time, the accuracy of cardiac THE was quantified both in silico and in vivo, specifically in a cohort including patients with ATTRwt. Multifrequency shear waves induced by portable drivers and encoded in standard ultrasound provide a rapid, low-cost echocardiographic contrast that aids the detection of cardiomyopathies such as amyloidosis.