10. Acoustic Models of Sound Production
417
normal fat to tendon. The lipids of the delphinid melon and lower jaw are,
however, chemically and acoustically distinct from other body fats and
blubber (Litchfield et al. 1973; Litchfield et al. 1975, 1979; Malins and
Varanasi 1975; Varanasi et al. 1975; Varanasi et al. 1982). It therefore makes
sense to correlate the velocities of all soft tissues except the melon and
lower jaw fats with their densities using a linear mapping from ordinary fat
to tendon for the characteristic mammalian soft tissue range. The unique
"acoustic fats" can then be incorporated by adding an extension from
normal fat down to the lower melon lipid density and velocity threshold.
Given the magnitude of the variations in mammalian soft tissue densities
and velocities, a two part linear mapping was deemed satisfactory for the
series of simulations reported here. The acoustic phenomena of interest to
this study result mainly from relatively short propagation paths (to wavelengths or less) through dermal layers, blubber, melon fat, muscle, and connective tissues. Because biologically reasonable variation of the velocity
mapping did not produce significantly different results, the velocities of
these tissues was presumed to be adequately represented by the current
model. This piecewise linear mapping of density versus velocity for the soft
tissue range is indicated by the solid line in Figure to.2b, demonstrating that
a good correspondence can be achieved between soft tissue velocity and
density. The current study is the first to have approximated the 3-D density
and velocity structures of all soft tissues including the melon and lower jaw
fats and to have accurately simulated 3-D acoustic propagation within those
structures. Diagrams of the resulting tissue density, velocity, and impedance
structure in a right parasagittal section through the core of the D. de/phis
melon can be found in Aroyan (1996). Previous analytical and ray tracing studies had assumed highly simplified two-dimensional tissue density and/or velocity models. (Compare, for example, Evans et al. 1964;
Dubrovskiy and Zaslavskiy 1975; Litchfield et al. 1979; Aroyan et al. 1992;
Au 1993).
Because the HU-to-density map has already been specified (Figure
10.2a), the empirical density-to-velocity map (Figure to.2b) then determines the soft tissue HU-to-velocity mapping. The HU-to-velocity mapping
FIGURE 10.2. (A) Plot of the linear mapping from CT values (in Hounsfield Units)
to tissue density used in the simulations. Tissue types corresponding to mapped CT
ranges have been indicated. (B) Plot of approximate values and ranges of density
and velocity for several types of normal terrestrial mammalian soft tissues and delphinid melon lipids measured at 37°C. Tendon A refers to propagation perpendicular to the tendon fiber bundles. Tendon B refers to propagation along the tendon
fiber axis. The solid line indicates the piecewise linear mapping of density to velocity used in the simulations over the range of soft tissues. (C) Plot of the mapping
from CT values (in Hounsfield Units) to tissue velocity used in the simulations. This
mapping (over the soft tissue range) is implied by the upper two figures. Tissue types
corresponding to mapped CT ranges have been indicated.
417
normal fat to tendon. The lipids of the delphinid melon and lower jaw are,
however, chemically and acoustically distinct from other body fats and
blubber (Litchfield et al. 1973; Litchfield et al. 1975, 1979; Malins and
Varanasi 1975; Varanasi et al. 1975; Varanasi et al. 1982). It therefore makes
sense to correlate the velocities of all soft tissues except the melon and
lower jaw fats with their densities using a linear mapping from ordinary fat
to tendon for the characteristic mammalian soft tissue range. The unique
"acoustic fats" can then be incorporated by adding an extension from
normal fat down to the lower melon lipid density and velocity threshold.
Given the magnitude of the variations in mammalian soft tissue densities
and velocities, a two part linear mapping was deemed satisfactory for the
series of simulations reported here. The acoustic phenomena of interest to
this study result mainly from relatively short propagation paths (to wavelengths or less) through dermal layers, blubber, melon fat, muscle, and connective tissues. Because biologically reasonable variation of the velocity
mapping did not produce significantly different results, the velocities of
these tissues was presumed to be adequately represented by the current
model. This piecewise linear mapping of density versus velocity for the soft
tissue range is indicated by the solid line in Figure to.2b, demonstrating that
a good correspondence can be achieved between soft tissue velocity and
density. The current study is the first to have approximated the 3-D density
and velocity structures of all soft tissues including the melon and lower jaw
fats and to have accurately simulated 3-D acoustic propagation within those
structures. Diagrams of the resulting tissue density, velocity, and impedance
structure in a right parasagittal section through the core of the D. de/phis
melon can be found in Aroyan (1996). Previous analytical and ray tracing studies had assumed highly simplified two-dimensional tissue density and/or velocity models. (Compare, for example, Evans et al. 1964;
Dubrovskiy and Zaslavskiy 1975; Litchfield et al. 1979; Aroyan et al. 1992;
Au 1993).
Because the HU-to-density map has already been specified (Figure
10.2a), the empirical density-to-velocity map (Figure to.2b) then determines the soft tissue HU-to-velocity mapping. The HU-to-velocity mapping
FIGURE 10.2. (A) Plot of the linear mapping from CT values (in Hounsfield Units)
to tissue density used in the simulations. Tissue types corresponding to mapped CT
ranges have been indicated. (B) Plot of approximate values and ranges of density
and velocity for several types of normal terrestrial mammalian soft tissues and delphinid melon lipids measured at 37°C. Tendon A refers to propagation perpendicular to the tendon fiber bundles. Tendon B refers to propagation along the tendon
fiber axis. The solid line indicates the piecewise linear mapping of density to velocity used in the simulations over the range of soft tissues. (C) Plot of the mapping
from CT values (in Hounsfield Units) to tissue velocity used in the simulations. This
mapping (over the soft tissue range) is implied by the upper two figures. Tissue types
corresponding to mapped CT ranges have been indicated.