NONPARAMETRIC SURFACE REGRESSION FOR STRAIN ESTIMATION
Joel E. Lindop, Graham M. Treece, Andrew H. Gee and Richard W. Prager
Ultrasonic strain imaging usually involves applying some form of smoothing filter to estimate gradients (corresponding to strains) after displacement measurements have been calculated from RF ultrasound data. All methods involve trade-offs between resolution and accuracy. Technical research often focuses on aspects of the RF signal processing, but differences in the smoothing method have a significant bearing on overall performance. We introduce a nonparametric regression method. Comparing with other linear-filtering techniques, this has advantages of higher image quality, greater versatility for practical applications, and (potentially) low computational cost. Its properties for strain imaging are examined through analysis and experiments. In addition to smoothing with uniform resolution, we demonstrate that nonparametric regression can be used to vary the resolution automatically based on indicators of data quality, thereby avoiding variation in the noise level within the image displayed. Interactions with RF signal processing parameters are also considered. This work indicates promising avenues for future research to improve the overall properties of practical strain imaging systems by taking a holistic approach to algorithm design.
If you have difficulty viewing files that end
which are gzip compressed, then you may be able to find
tools to uncompress them at the gzip
If you have difficulty viewing files that are in PostScript, (ending
'.ps.gz'), then you may be able to
find tools to view them at
We have attempted to provide automatically generated PDF copies of documents for which only PostScript versions have previously been available. These are clearly marked in the database - due to the nature of the automatic conversion process, they are likely to be badly aliased when viewed at default resolution on screen by acroread.