We applied a newly developed wavelet-based method [Wu et al. 2000] to track objects in ultrasonic image sequences. The dense motion vectors between two successive images were represented by a linear combination of coarse-to-fine basis functions developed recently by Cai and Wang [Cai and Wang, 1996]. The coarser-scale basis function has larger support while the finer-scale basis function has smaller support. By minimizing the SSD (Sum of Squared Differences) iteratively, coarse-to-fine basis functions were used to match large-to-small full-resolution regions concurrently in recovering the motion vectors and thus tracking the target. Experimental results of tracking fetal heart in the ultrasonic image sequences were demonstrated. We further applied the PCA (Principal Component Analysis) to extract the deformation of tracked heart contour. The significant deformation along with other physiological parameters of fetuses can be employed in the application of predicting SIDS (Sudden Infant Death Syndrome) in future studies.
|Number of pages||8|
|Journal||Journal of Medical and Biological Engineering|
|State||Published - Jun 2002|
- Coarse-to-fine wavelet
- Object tracking
- Sum of squared difference (SSD)
- Ultrasonic images