There is also the possibility to export the characteristic
vectors obtained in .xls format, in order to be processed and
analyzed using other Data Mining tools.
A database of ultrasound images may be built using this
procedure. Our database contains a total of 67 images for a
total of 21 persons. For each person there are a number of 2
to 5 images. The database images are indexed according to
the signature on each image they present. There have been
two types of tests:
(A) A total of five people, whose ultrasound images are
found in the database, were selected. These people have
every new ultrasound image. Several persons whose images
are similar were searched using this set of images in the
database. The results are presented in Table 1.
The results (similarity percent) show that it is possible to
find a similar image according to the information present in
the image.
(B) A total of five people, whose scans are not in the
database, were selected. These people have every ultrasound
images. For these people, the medical diagnosis is known.
Several persons whose images are similar were searched
using this set of images in the database. The results are
presented in Table 2.
The similarity percent in the last row indicates the
measure in which the known diagnosis for these new people
corresponds to that of similar individuals found. The results
obtained show that it is possible to obtain a proper diagnosis
of an unknown image by comparing the information present
in an ultrasound image.