Module 3-4
Limitations and Real-World Considerations
The μ = ρc² relationship rests on several simplifying assumptions: the tissue is linearly elastic, isotropic, homogeneous, and geometrically unbounded. None of these holds exactly for the myocardium. The following considerations describe where each assumption breaks down and its implications for interpreting SWE measurements.
Viscoelasticity
Press your hand into memory foam quickly versus slowly: a fast push meets more resistance than a slow one. Myocardial tissue behaves similarly — its response depends on how fast it's deformed, which means different frequency components of a shear wave travel at different speeds.
The μ = ρc² derivation assumes purely elastic behavior. In this model, energy is fully stored and returned with no loss. However, the myocardium is viscoelastic: its mechanical response includes both elastic and viscous components, which means it resists more when stretched faster and recovers slowly and dissipates energy over time. In a purely elastic material, all frequency components of a shear wave travel at the same speed. In a viscoelastic material, higher-frequency components travel faster than lower-frequency ones, which is known as shear wave dispersion.
Because ARF-induced waves and natural valve-closure waves arise from different mechanical events, they carry different frequency content. In a dispersive medium such as the myocardium, this produces systematic differences in velocity estimates between the two methods, even in the same tissue at the same cardiac phase.
ARF-induced wave
Natural (valve-closure) wave
animation: ARF & natural valve-closure waves, shown on separate rows for clarity — both depart at the same moment
velocity vs. frequency — dispersion curve (solid) vs. non-dispersive reference (dashed)
In a purely elastic tissue (viscoelasticity: none), the dispersion curve coincides with the reference line and every frequency travels at the same speed, so the ARF and natural points land at the same height and the two waveforms above stay locked together. As viscoelasticity increases, higher-frequency components travel faster than lower-frequency ones, so the curve bows upward. Because ARF-induced and natural valve-closure waves carry different frequency content, this same dispersion produces a systematic velocity gap between the two methods, in the same tissue at the same cardiac phase. Note: Frequencies, velocities, and timing are illustrative, not calibrated to a specific measurement.
Anisotropy
The conversion of shear wave velocity to a shear modulus assumes isotropic behavior: identical mechanical properties in all directions. The LV myocardium is structurally anisotropic. Cardiomyocyte fibers are arranged in a transmural helix, with the fiber orientation rotating continuously from the epicardium to the endocardium relative to the circumferential direction. Shear wave velocity varies with the angle between the propagation direction and the local fiber orientation, with waves generally propagating faster along the fiber axis than across it. A single velocity measurement along one propagation direction does not capture this directional dependence, and derived modulus values should be interpreted with the acquisition geometry in mind.
The isotropic panel's speed stays fixed at 1.00 regardless of either slider — direction and depth have no effect on it. The anisotropic panel's speed changes with both: it is highest when the propagation angle aligns with the local fiber direction and lowest across it, and that fiber direction itself rotates as the transmural depth slider moves. Note: this figure is schematic. The anisotropy ratio and the endocardial/epicardial fiber angles are illustrative placeholders, not values drawn from a specific measurement.
Guided waves and wall geometry
We also assume that wave propagation is in an unbounded medium. The myocardial wall is a finite-thickness structure bounded by the endocardium and epicardium. When shear wavelength is comparable to wall thickness, the bounding surfaces begin to constrain wave propagation, and velocity becomes dependent on wall geometry as well as tissue mechanical properties. Velocity is not uniquely determined by shear modulus alone, and applying μ = ρc² without accounting for wall thickness introduces error into stiffness estimates. This effect is most pronounced in pediatric imaging, where LV wall thickness may be only a few millimeters. Wall thickness should therefore be documented alongside velocity measurements as a relevant co-variable in pediatric myocardial SWE.
Wall
wall cross-section — wavelength fixed at 5 mm; wall thickness follows the slider above
apparent velocity vs. wall thickness / wavelength — dashed line marks the true bulk shear velocity
At this wall thickness, the apparent velocity is 32.9% below the true bulk shear velocity — applying μ = ρc² directly here would underestimate the tissue's true stiffness. As wall thickness grows relative to the wavelength, the boundaries matter less and the apparent velocity converges toward the true value. This is why wall thickness should be documented alongside velocity in pediatric myocardial SWE, where LV wall thickness can approach the shear wavelength itself. Note: The curve's shape is illustrative — not a full Rayleigh–Lamb solution — and is not calibrated to a specific measurement.
Attenuation and reflections
Shear wave amplitude decreases with propagation distance due to geometric spreading, viscous absorption, and scattering. At greater distances from the push site, displacement amplitude may fall below reliable detection thresholds, limiting the spatial extent over which velocity estimates can be obtained. Reflections from tissue boundaries, including the endocardium, pericardium, and structural discontinuities, can produce secondary wave fronts that interfere with the primary propagating wave and complicate velocity estimation.
Summary
The shear modulus model assumes elastic, isotropic, and unbounded tissue.
- In the myocardium, viscoelasticity makes wave velocity frequency dependent.
- Anisotropic fiber architecture means wave velocity varies with propagation direction.
- In thin walls, the bounding surfaces constrain wave propagation, making wall thickness a relevant co-variable — an effect most pronounced in pediatric imaging.
- Attenuation and boundary reflections set practical limits on measurement range and reliability.