运用定量形态学测量法检定原发性脊椎骨折

The application of auantitative morphometry for assessment of prevalent vertebral fracuture: comparison with semiquantitative technique

为比较半定量评定法（ＶｉｓｕａｌＳｅｍｉｑｕａｎｔｉｔａｔｉｖｅＡｓｓｅｓｓｍｅｎｔ，ＳＱ）与定量形态学测量法（ＱｕａｎｔｉｔａｔｉｖｅＭｏｒｐｈｏｍｅｔｒｉｃＡｓｓｅｓｓｍｅｎｔ，ＱＭ）对脊椎骨折的检测结果，随机抽样４００名用ＤＸＡ测量有骨量减少的绝经后妇女。她们的侧位标准胸腰椎平片均由三位经培训的放射科医师评阅（使用同一骨折判定标准）。并由一位放射科医师运用“六点法”测量每一椎体的前、中、后高度（Ｔ４－Ｌ４）。ＱＭ法的骨折阈值分别定为低于正常人椎体高度比例的平均值的２．０，２．５，３．０，３．５和４．０个标准差（即２．０ｓ，２．５ｓ，３．０ｓ，３．５ｓ，４．０ｓ），分析两种方法所检出的骨折椎体，再由第４位放射科医师将由两种诊断方法存在差异的椎体重新分析以确定此椎体是否有骨折，此结果作为“诊断标准”（ＣＲ）。各放射科医师间的Ｋａｐｐａ值为０．８０～０．８１，各放射科医师与ＣＲ间的Ｋａｐｐａ值为０．８４～０．８７，敏感性为８５．１７～９１．７５％，特异性为９７．１９～９８．０６％。ＱＭ法的各骨折阈值与ＣＲ间的Ｋａｐｐａ值为０．５４～０．７５，其中以２．５ｓ阈值与ＣＲ结果最为符合。其Ｋａｐｐａ值为０．７５，敏?

To compare visual semiquantitative and quantitative rnorphometric methods of vertebral fracture assessment, radiographs of the lateral thoracic and lumbar spine of 400 postmenopausal women with osteopenia, as defined by spinal DXA measurements, were evaluated independently by three trained radiologists. All radiographs were digitized by a radiologist and the presence of vertebral fractures was determined using standard deviation (SD)reductions of vertebral height ratios of 2.0, 2.5, 3.0, 3.5and 4.0 as cutoff criteria. The results from a joint consensus reading (CR) of the radiographs including a fourth experienced radiologist was compared to the results from the individual readings and quantitative morphometry. A kappa statistic was applied to assess the agreement among all readers, morphometry and the consensus reading. Kappa values among the three radiologists ranged from 0.80to0.81; between the radiologists'readings and the CR, kappa values ranged from 0.84 to 0.87. Sensitivities and specificities of the three radiologists with respect to the consensus reading ranged from 85.17%to91.75% and from 97.19% to 98.06%, respectively. The kappa values for the agreement between morphometry and the CR ranged from 0. 54 to 0. 75 using the above cutoff criteria. Compared to the CR ,the kappa score for morphometry at 2. 5 SD was 0. 75, with a sensitivity and specificity of 70. 23% and 98.76%, respectively. Our results indicate that the concordance among trained radiologists for semiquantitative assessment of vertebral fractures is excellent and is higher than that between semiquantitative and quantitative methods regardless of the cutoff criterion used. Using a reduction of 2. 5 SD in vertebral height ratios from normals as a cutoff criterion for morphometry provided the highest concordance with the consensus reading for assessing prevalent fractures.