目的 探讨能谱CT在膝关节急性外伤性骨髓水肿中的应用价值.方法 选取行能谱CT和MRI检查的急性膝关节外伤患者32例.能谱CT扫描数据传输到后处理工作站AW4.7,以MRI作为参考标准,由2位诊断医师分别利用能谱水-钙(Ca)、水-羟基磷灰石(Hap)彩色编码图像评估骨髓水肿程度,比较两种方法评估骨髓水肿差异有无统计学意义;测量水肿区域的水-Ca、水-Hap值,评价差异有无统计学意义,并利用ROC曲线分析两者在骨髓水肿中的诊断效能.结果 32例膝关节外伤,MRI显示骨髓水肿区域146处,能谱CT中:水-Ca彩色编码图像共获得135处水肿区域,灵敏度为95.6％,特异度为98.8％,阳性预测值为96.3％,阴性预测值为98.5％;水-Hap彩色编码图像共获得128处水肿区域,灵敏度为90.4％,特异度为98.8％,阳性预测值为96.1％,阴性预测值为96.9％;MRI、水-Ca、水-Hap三者之间差异无统计学意义(P＞0.05).ROC分析显示在水-Ca中,当相对水含量值为1 020.02 mg/cm3时,能谱CT对骨髓水肿的诊断效能最高,灵敏度99.3％,特异度74.1％,AUC为0.93;在水-Hap中,当相对水含量值为953.99 mg/cm3时,能谱CT对骨髓水肿的诊断效能最高,灵敏度88.1％,特异度93.3％,AUC为0.96,差异有统计学意义(P＜0.001).结论 能谱CT的水-Ca、水-Hap均可以清晰显示急性膝关节外伤患者的骨髓水肿区域,并精准定量评估:MRI、水-Ca、水-Hap彩色编码图像显示骨髓水肿区域差异无统计学意义;利用ROC曲线分析显示水肿区域的水-Ca、水-Hap值在骨髓水肿中的诊断效能差异有统计学意义.

Early and accurate assessment of lumbar intervertebral disc degeneration (IVDD) is very important to therapeutic strategy. This study aims to correlate and compare the performances of T1ρ, T2 and T2* mapping for Pfirrmann grades and morphologic changes in the IVDD. This prospective study included 39 subjects with 195 lumbar discs. T1ρ, T2 and T2* mapping were performed, and T1ρ, T2 and T2* values of nucleus pulposus (NP), and anterior and posterior annulus fibrosus were measured. IVDD was assessed with Pfirrmann grading and morphologic changes (normal, bulging, herniation and annular fissure). The performances of T1ρ, T2 and T2* relaxation times were compared for detecting early (Pfirrmann grade II-III) and advanced degeneration (Pfirrmann grade IV–V), as well as for morphologic changes. T2 relaxation times was strongly corelated with T1ρ and T2* relaxation times. Areas under the curves (AUCs) of T1ρ, T2 and T2* relaxation times of NP were 0.70, 0.87 and 0.80 for early degeneration, and 0.91, 0.95 and 0.82 for advanced degeneration, respectively. AUCs of T1ρ, T2 and T2* relaxation times of NP were 0.78, 0.83 and 0.64 for bulging discs, 0.87, 0.89 and 0.69 for herniated discs, and 0.79, 0.82 and 0.69 for annular tearing, respectively. The AUC of T2 relaxation time was significantly higher than those of T1ρ relaxation times (both P < 0.01) for early IVDD, and the AUCs of T1ρ and T2 relaxation times for assessing advanced degeneration and morphologic changes were similar (P > 0.05) but significantly higher than that of T2*relaxation time (P < 0.01). T2 mapping performed better than T1ρ mapping for the detection of early IVDD. T1ρ and T2 mapping performed similarly but better than T2* mapping for advanced degeneration and morphologic changes of IVDD.