The Ultrasound Visualization Pipeline - Scientific Visualization: Uncertainty, Multifield, Biomedical, and Scalable Visualization - page 278

Fig. 24.6 The parameter I thresh determines which echo intensity values to render transparently.
A user control with immediate feedback, indicating transparent regions in pink, is essential
a default OTF providing reasonable visualization in the majority of cases, and hardly
ever touch the OTF control panel.
Therefore, there is a need for automatic determination of an optimal OTF for
every single acquisition. Due to the distinct characteristics and the real-time nature
of ultrasound imaging, most conventional approaches for transfer function design
have proven inadequate or require substantial modification in order to be applicable to
ultrasound volume imaging. Among the most important advances in transfer function
design for CT data is the work by Kindlmann et al. [ 34 ] and subsequent work by
Kniss et al. [ 36 ], introducing the concept of histogram volumes for semi-automatic
generation of OTFs for datasets where the regions of interest are boundaries between
materials of relatively constant value. In [ 30 ], von Jan et al. adapt this approach
to ultrasound data and apply it successfully to 3D freehand acquired volumes of
hyperechoic structures.
Hönigmann et al. suggest an approach dedicated to the specific problem of ren-
dering hyperechoic structures embedded in hypoechoic fluid [ 28 ]. By analyzing so
called tube cores , they yield an estimate for the position of the most prominent tissue
transition in the rendering direction. Voxel intensities prior to and at the detected in-
terface steer the extent of modification of an initial, parabolic OTF in a multiplicative
way. In a subsequent publication the authors assess the temporal coherence of the
tube core method and conclude that it is sufficiently efficient and robust for online-
computation of OTFs for an entire sequence of acquired volumes, if smoothing in
the temporal domain is employed [ 54 ].