Bone Mineral Density Analysis with Thalassemia X-ray Image Processing Using GLCM and First-Order Statistic Methods

Thalassemia is a high-risk genetic disorder that can lead to bone complications, including osteoporosis, due to chronic iron overload from blood transfusions. BMD examinations that commonly use DXA still face a number of limitations, especially in terms of accessibility and anatomical variations. This study offers another approach in quantitatively estimating BMD values through X-ray image processing, utilizing GLCM-based first-order statistical and texture feature extraction. The analysis process was carried out in stages, starting from image preprocessing, feature extraction, and testing correlation relationships to building a linear regression model that was validated using a prediction error measure. Among all the features analyzed, kurtosis in the non-thalassemia data group showed the strongest linear relationship with BMD values, evidenced by a Pearson correlation value of r = 0.913 and a coefficient of determination (R<sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sup>) of 0.841. This indicates that kurtosis has enormous potential to be used as an imaging indicator in estimating BMD. Specifically, homogeneity 90° showed a very strong correlation with BMD in the thalassemia group (r = 0.93), demonstrating its potential as a specific imaging biomarker for this group. This approach also opens up opportunities for a more cost- and time-effective method of measuring bone density, especially in areas with limited diagnostic tools. Hopefully, these results can contribute to the development of a more affordable osteoporosis screening system, especially for high-risk populations such as people with thalassemia.