Ju Hwan Lee , Sung Yun Park , Jae Hoon , Sung Min Kim

The KIPS Transactions:PartB , Vol. 19, No. 2, pp. 99-106, Apr. 2012
10.3745/KIPSTB.2012.19.2.99,   PDF Download:

Abstract

The purpose of this study was to determine the evaluation parameters`` osteoporosis predictability in accordance with measuring regions by analyzing the correlations between bone mineral density and trabecular patterns derived from different measuring regions. Experimental subjects were a total of 40 female patients after menopause aged over 40 years, and were classified into 20 control and 20 osteoporotic groups according to the T-score. Bone mineral density was measured on femoral neck, trochanter and ward``s triangle by DEXA (Dual Energy X-ray Absorptiometry). We designated ROI (Region of Interest) with 50 x 50 pixel size on each measuring regions, and extracted trabecular patterns by using existing image processing method. We also selected a total of eight evaluation parameters that are categorized into structural (mean gray level, area, perimeter, thickness and terminal distance), skeletonized parameters (number, length) and fractal dimension. As a result, it was observed that area, perimeter, thickness, terminal distance, number, length and fractal dimension reflected the bone mineral density with high statistical validity (p<0.003). We also confirmed that the evaluation parameters could predict the osteoporosis more efficiently.

Statistics
Cite this article