Characterization of bone quality using computer‐extracted radiographic features

Abstract

Both bone mineral density (BMD) and trabecular structure are important determinates of bone mechanical properties. However, neither BMD or trabecular structural features can completely explain the variations in bone mechanical properties. In this study, we combine BMD and bone structural features to characterize bone mechanical behavior. Radiographs were obtained from 34 femoral neck specimens excised during total hip arthroplasties. Each neck radiograph was digitized and a region of interest (ROI) was selected from the medial side of the femoral neck. Textural features, the global Minkoswski dimension and trabecular orientation, were extracted from each ROI image using Minkowski dimension analysis. The BMD of each specimen was measured using dual-energy x-ray absorptiometry (DXA) and subsequently normalized by bone size as measured from a standard pelvis radiograph. Mechanical testing was performed on the trabecular bone cubes machined from each femoral neck to yield bone mechanical properties. Multiple regression was performed to select the best features to predict bone mechanical properties. The results suggest that, using multiple predictors including normalized BMD structural features, and patient age, the coefficients of determination (R2) improved over the use of BMD alone. For bone strength, the R2 was improved from 0.24 using conventional BMD to 0.48 using a four-predictor model. Similar results were obtained in the prediction of Young's modulus, i.e., the R2 was improved from 0.25 to 0.55 in going from the model using conventional BMD to a four-predictor model. This study demonstrates the contributions of normalized BMD, structural features, and age to bone mechanical properties, and suggests a potential method for the noninvasive evaluation of bone mechanical properties.

Development and applicability of a quality control phantom for dental cone‐beam CT

“…Bone analysis methods do not, however, necessarily rely on absolute or relative voxel values. ( 21 ) The results demonstrate that a CT number correlation analysis should not be part of a QC procedure for dental CBCT, as many manufacturers do not claim to have an accurate correlation with CT numbers for their devices. Nevertheless, these devices can still be used in clinical practice for analysis of bone quality.…”

Variability of dental cone beam CT grey values for density estimations

“…Providing that large sample sizes are used, varying CBCT devices are included and a worthy gold standard is available, dedicated and validated bone quality parameters could be determined. Jiang et al [41] investigated the combination of BMD and structural analysis to predict mechanical properties of the bone, finding an improvement in the predictive value when combining multiple parameters. Possibly, a prediction model for CBCT with multiple variables will offer the most accurate correlation with mechanical properties of the bone.…”

Computerized analysis of mammographic parenchymal patterns using fractal analysis

“…Many physical systems in nature are suitable to be described by fractal dimensions to show both self similarity and the overall roughness at multiple scales 5 . Several studies reported that fractal dimension has been useful in characterizing complex biological systems which are similar to mammographic patterns [6][7]10 . All these suggest that fractal based computerized radiographic markers may be a useful descriptor to characterize mammographic parenchymal patterns.…”