Fatal error: Call to undefined function post_password_required() in /home/agusza/public_html/nikki/wp-content/themes/avelius/comments.php on line 9
Categories
- Agama (28)
- Aswaja (64)
- Bahasa Indonesia (27)
- Blogroll (1)
- elearning (14)
- English (7)
- ITS (19)
- Kampus (18)
- Komputer (13)
- Ngaji (2)
- Pattern Recognition (4)
- Sekolah (11)
- Uncategorized (39)
Archives
- February 2010
- January 2010
- December 2009
- October 2009
- September 2009
- July 2009
- June 2009
- May 2009
- April 2009
- March 2009
- February 2009
- January 2009
- December 2008
- November 2008
- October 2008
- September 2008
- August 2008
- July 2008
- June 2008
- May 2008
- April 2008
- March 2008
- February 2008
- January 2008
- March 2007
- March 2006
Fatal error: Call to undefined function post_password_required() in /home/agusza/public_html/nikki/wp-content/themes/avelius/comments.php on line 9
Kuliah Tamu
http://www.math.kyushu-u.ac.jp/english/calendar/seminar_view.php?kind=&S_Category=&S_Language=&S_Page=&S_View=&word=&page=&B_Code=886
to assess health status of oral and dental structures. As osteoporosis
is a systemic skeletal disease, it also affects bone density and
structure of the jaws. Recently, considerable attention has been paid to
low skeletal bone mass screening based on mandibular parameters. The
results of dental panoramic radiographs signs are promising in
osteoporosis risk prediction. It was found that the computer-aided
diagnosis scheme statistically significantly improved diagnostic
accuracy, particularly for radiologists with less experience. We
previously developed a computer-aided system for measuring mandibular
cortical width on dental panoramic radiographs. We have compared
skeletal BMD and cortical width measured by the system. These
correlations were similar to those between skeletal BMD and cortical
width measured manually. Sensitivity and specificity for identifying
postmenopausal women with low spinal BMD by our computer-aided system
were about 88.0% and 58.7%, respectively. Since the measured parameter
does not have highly sensitivity and specificity to diagnose patients
with osteoporosis, more feature extractions are needed. Reduced
mandibular cortical and alveolar bone mineral density, decreased
alveolar bone height, increased porosity of mandibular alveolar and
inferior cortical bone, erosion and reduced width of the inferior border
of the mandible, and altered trabecular patterns in maxillary and
mandibular bone have been demonstrated in osteoporotic patients.
Interest in this image processing research is due to the unevenly
illumination, low contrast, and domination of dark color which are seen
on the image. Since the boundary of each object was not sharp, we need
robust algorithms to separate bone from remaining objects. Thresholding,
which assigns a pixel to one class if its gray level is greater than a
specified threshold and otherwise assigns it to the other class, is the
most common method for separating objects from background. We have
developed algorithms for thresholding the images based on discriminant
analysis, fuzzy logic type I and II, and multistage adaptive algorithm.
We applied average filter as low-pass filter to the original image and
subtracted this low-frequency image from the original image to obtain
the image containing only high frequencies. Morphological open and
closed operations are important for enhancement processes by joining
small patches that spread out along the bone.
Another work has been done for segmenting the cortical bone using
combination of watershed and active contour GGVF snake. The main
advantage is the initialization area can be far enough from the object,
because active contour may trace the segments resulted by watershed
method. For measuring the alveolar height, we combined the multistage
adaptive thresholding and genetic algorithm to obtain clear upper and
lower boundary along the mandible, because solid and compact edges are
needed to measure it appropriately. In the sense of trabecular bone,
there has been several works for analyzing tissues structure inside the
mandible to know altered trabecular patterns. We segmented the bone
based on the statistical textural information which are from first and
second-order. We also generated random graph with erdos royi methods
for measuring the degree and cluster coefficient to know the difference
between subjects which identified as osteoporosis and normal. Another
work for analyzing trabecular structure is by obtaining linear
structure, orientation, and scale such that each trabeculae can be
segmented well. Future work will develope further image analysis for
extracting features of the mandible to improve the diagnostic accuracy
of our methodology.
http://www.mis.hiroshima-u.ac.jp/missemi2008.html