Limitations of quantitative bone‐mass measurements using assessments of first‐order statistics of grey‐level histograms in plain radiographs

Abstract

Purpose: To evaluate the quality assurance of image‐processing techniques in plain radiographs of skeletal structures. Material and Methods: Twenty‐two patients were studied, each with one osteolytic metastasis. Accuracy and precision of tube voltage and timer were confirmed. The mean value of grey‐level histograms in plain radiographs (MVGLHs) was assessed. The deviation was monitored after five sets of sequential X‐rays retaining the same settings for each radiograph. Results: Deviation was significantly higher in anatomical areas of thorax (21.2%) and abdomen (42.4%), while the consistency of MVGLH for weight‐bearing bones was satisfactory with a maximum deviation of 2.9% ( P<0.001, Kruskal‐Wallis test). Conclusion: Assessment of MVGLH in plain radiographs is a reliable method for the extremities and generally for regions without superimposed movable tissues.