Graphics Reference
In-Depth Information
Figure
.
.
Segmentation results are displayed for a sequence of MR images of myocardium in the
frequency domain. he top three images display classification results and the bottom three images show
prediction results
Table
.
.
Classification error rates for a sequence of MR imageswhere Obj represents the object, Bg
denotes the background, and Total refers to the average error of the whole image
Frame
Frame
Frame
Feature Obj
Bg
Total
Obj
Bg
Total
Obj
Bg
Total
Space
.
.
.
.
.
.
.
.
.
FFT
.
.
.
.
.
.
.
.
.
Gabor
.
.
.
.
.
.
.
.
.
outer walls of the let ventricle endocardium. he training set can be obtained from
the first three images by a medical expert. Using the proposed procedure, the seg-
mentation results in the three feature domains shown in Figs.
.
,
.
, and
.
. he
classification error rates (derived based on the target boundaries drawn by a doctor)
arereportedinTables
.
and
.
.hus,DSIRsuccessfully performs thesegmentation
of this sequence of MR images. More studies are reported in Wu and Lu (
).
Conclusion and Discussion
2.5
hisarticle introduces and discussesseveral studies on the reconstruction, visualiza-
tion and analysis of medical images. In particular, computational statistical methods
are used as important tools in the analysis of PET, ultrasound and magnetic reso-
