Research: Quantitative Ultrasound Imaging
This research is concerned with methods of improving diagnostic ultrasound. Currently,
ultrasound scanners operate in a very qualitative manner. Ultrasound images are
representations of the relative echo amplitude, without any consideration for
comparison with an absolute scale. The images that are produces are wonderfully
detailed and allow radiologists and physicians an amazing view. Recently, interest
has been rekindled in "quantitative ultrasound imaging", in which images would
represent numerical values of acoustical properties of tissue. This may offer the
following advantages
- Enhanced diagnosis: It is known that certain diseases, inlcuding cirrhosis
of the liver and some malignant cancers, change the values of acoustical properties
in prescribed ways. Directly measuring these properties may enhance clinical
evaluation of these diseases
- Stability of images: As ultrasonographers and scanners change, the images from
current scanners may also change. Since the images only reflect relative amplitudes,
there is limited ability to ensure that an image taken today with machine A will
be similar to one taken tomorrow by machine B. Quantitative ultrasound imaging
circumvents this difficulty by producing images that directly reflect underlying
physical properties of tissue.
- Reduction of artifacts: The amplitude of an ultrasound echo is affected by
two physical properties: attenuation, largely the absorption of ultrasound energy
by tissue, and backscatter, the scattering of sound energy back to the source. It is
possible that a region with high attenuation will absorb so much ultrasound that
regions below it will appear black. This is like a shadow where the light has been
blocked. Because of this, little ultrasound energy is available to interact with
areas below this region and thus there is little echo. This mimics a similar condition
in which an area may have low backscatter and thus not produce much echo. By separating
images into an attenuation image and a backscatter image (both quantitatively determined)
we can prevent this "confusion" of the two possibilities.