Methods for automated segmentation of trabecular bone structure

The precise segmentation of bone regions is important in applications where measurements are taken from the extracted tissues. In high quality biomedical images, these extracted regions can be processed to study the underlying biology, like the relationship between bone structure and genetics. In studying the relationship between trabecular tissue growth patterns and the genes governing that particular development in mice limbs, it is important to ensure that segmentation is accurate. After obtaining the region of interest, image analysis techniques can be non-invasively applied to make any desired calculations. Fractal analysis has been used to measure structure related factors in trabecular bone. Since conventional methods could result in substantial loss of trabecular bone volume, methods for automated segmentation are compared. The first method performs automated segmentation using thresholding and morphological operators, while the second uses region growing. A comparison between two segmentation approaches is done. According to the fractal analysis, the first method outperforms the second in the segmentation precision.