| Summary: | Using ultrasound for tool navigation in an orthopaedic surgery requires bone localization in ultrasound, a procedure which remains challenging despite encouraging advances in its practice. Current methods, e.g., the local image phase-based
feature analysis, have shown promising results but continue to rely on delicate
parameter selection processes and to be prone to error at confounding soft tissue
interfaces of similar appearance to bone interfaces. In addition, the 3-dimensional
phase-features-based automatic bone segmentation method is found to be time
inefficient (at ~2 minutes).
We have proposed a different approach to bone segmentation by combining
ultrasound strain imaging and envelope power detection methods. After an estimation of the strain and envelope power maps, we subsequently fused these
maps into a single combined map that corresponds robustly to the actual bone
boundaries. This study has achieved a marked reduction in false positive bone
responses at the soft tissue interfaces. We also incorporated the depth-dependent
cumulative power of the envelope into the elastographic data as well as incorporated an echo de-correlation measurement-based weight to fuse the strain and
envelope map. We also employed a data driven scheme to detect the presence of
any bone discontinuity in the scanned ultrasound image and introduced a multivariate non-parametric Gaussian mixture regression to be used over the maximum
intensity points of the fused map. Finally, we developed a simple yet effective
means to perform 3-dimensional bone surface extractions using a surface growing
approach that is seeded from the 2-dimensional bone contours.
We employed mean absolute error calculations between the actual and estimated bone boundaries to show the extent of the false positives created; our
methods showed an average improvement in the mean absolute error of 20% on
both the 2- and 3-dimensional finite-element-models, and of 18% and 23%, respectively, on the 2- and 3-dimensional experimental phantom data, when compared with that of the local phase-based methods. Validation on the 2- and
3-dimensional clinical in vivo data also demonstrates, respectively, an average
improvement in the mean absolute fitting error of 55% and an 18 times improvement in the computation time. === Applied Science, Faculty of === Graduate
|
|---|
