采用欧洲腰椎体模对多中心腰椎定量CT的精密度和准确度评估

Assessment of Precision and Accuracy of Spinal QCT in a Multi-centre Clinical Research with a European Spine Phantom

目的采用欧洲腰椎体模(ESP)作为标准评估多中心腰椎定量CT(QCT)骨密度测量的重复性和准确性误差。材料与方法2个中心(3台CT分别是东芝64排CT、东芝16排CT及GE64排CT)按照常规腰椎定量CT骨密度测量条件和方法对欧洲腰椎体模(ESP)进行重复测量各10次;另外4个中心(CT分别是东芝64排、东芝16排、西门子64排、GE双排)分别用欧洲体模(ESP)扫描测量一次。结果 3台QCT(CT顺序依次是东芝64排,东芝16排,GE64排)对欧洲体模测得的相对精密度CV分别是0.512%、0.523%、0.514%。7台QCT机器间的精密度误差CV分别是L14.213%、L22.752%、L34.786%。7台CT的准确度变化范围和平均值(括号因数值)分别为L11.20%～8.79%(5.1);L2-0.25%～9.15%(3.8);L3-7.03%～5.69%(-0.5)。结论参与研究的多中心QCT系统的机器内精密度、机器间精密度及准确度较好,结果准确可靠。

Purpose To evaluate the reproducibility and accuracy of spinal QCT in a multi-centre clinical research using a European Spine Phantom. Material and Methods ESP was scanned 10 times using routine technique without repositioning the phantom on 3 CT scanners （64-row Toshiba, 16-row Toshiba and 64-row GE CT scanner） in two centers. The ESP was scanned once following the same protocols in another four centers using 64-row Toshiba, 16-row Toshiba, 64-row Siemens and 2-row GE CT scanner respectively. Results The coefficient of variation of the first three QCT systems was 0.512%, 0.523% and 0.514% respectively for integral phantom. The inter-modality precision errors of seven QCT systems were 4.213%, 2.275%, and 4.786% in L1, L2 and L3 respectively. The range （and mean） of accuracy error of seven QCT systems were respectively 1.20% to 8.79% （5.1）, -0.25% to 9.15% （3.8） and -7.03% to 5.69% （-0.5） in L1, 1-2 and L3. Conclusion Spinal QCT shows excellent precision, intermodality precision and accuracy in this multi-centre clinical research.