A MODEL-BASED APPROACH TO 3D FREEHAND ULTRASOUND IMAGING
Mike Syn and Richard Prager
An iteratively refined shape recovery procedure for use in 3D freehand ultrasound imaging is presented. This procedure utilises models of both 3D shape and shape variation, by using vibration modes from finite element analysis in the initial shape model, and updating the model using principal components analysis (PCA). The difficulty in applying common PCA schemes, which require prior segmentation of training examples, is thus avoided. This is very important, since manual segmentation in 3D ultrasound is impractical for a reasonably sized training set, particular it has to be performed for each new class of shapes. A scale and rotation invariant shape registration procedure is also presented. Such a procedure allows the quantitative comparison of segmented shapes from ultrasound with other modalities for the purposes of validation. It also allows us to address the issue of automated specification of homologous landmarks, by registering a densely specified example which has just been segmented with the current model. The issues of model construction and volume measurement are also discussed.
Keywords: shape models, vibration modes, principal components, automated registration, model construction, volume measurement.
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.