Abstract：
OBJECTIVE: The purpose of this study was to determine whether CT attenuation thresholds can be used to distinguish untreated osteoblastic metastases from enostoses.
METHODS: The study group comprised 62 patients with 279 sclerotic bone lesions found at CT (126 enostoses in 37 patients and 153 metastases in 25 patients). The cause of sclerotic lesions was assessed histologically or by clinical and imaging follow-up. None of the patients had undergone prior treatment for the metastases. The mean and maximum attenuation were measured in Hounsfield units. ROC analysis was performed to determine sensitivity, specificity, AUC, 95% CIs, and cutoff values of CT attenuation to differentiate metastases from enostoses. Interreader reproducibility was assessed using an intraclass correlation coefficient with 95% CI.
RESULTS: The mean and maximum CT attenuation values of enostoses were 1190 ± 239 HU and 1323 ± 234 HU, respectively, and those of osteoblastic metastases were 654 ± 176 HU and 787 ± 194 HU, respectively. Using a cutoff of 885 HU for mean attenuation, the AUC was 0.982, sensitivity was 95%, and specificity was 96%. Using a cutoff of 1060 HU for maximum CT attenuation, the AUC was 0.976, sensitivity was 95%, and specificity was 96%. The mean attenuation intraclass correlation coefficient was 0.987 for enostoses and 0.81 for metastases. The maximum attenuation intraclass correlation coefficient was 0.814 for enostoses and 0.980 for metastases.
CONCLUSIONS: CT attenuation measurements can be used to distinguish untreated osteoblastic metastases from enostoses. A mean attenuation of 885 HU and a maximum attenuation of 1060 HU provide reliable thresholds below which a metastatic lesion is the favored diagnosis.
METHODS: The study group comprised 62 patients with 279 sclerotic bone lesions found at CT (126 enostoses in 37 patients and 153 metastases in 25 patients). The cause of sclerotic lesions was assessed histologically or by clinical and imaging follow-up. None of the patients had undergone prior treatment for the metastases. The mean and maximum attenuation were measured in Hounsfield units. ROC analysis was performed to determine sensitivity, specificity, AUC, 95% CIs, and cutoff values of CT attenuation to differentiate metastases from enostoses. Interreader reproducibility was assessed using an intraclass correlation coefficient with 95% CI.
RESULTS: The mean and maximum CT attenuation values of enostoses were 1190 ± 239 HU and 1323 ± 234 HU, respectively, and those of osteoblastic metastases were 654 ± 176 HU and 787 ± 194 HU, respectively. Using a cutoff of 885 HU for mean attenuation, the AUC was 0.982, sensitivity was 95%, and specificity was 96%. Using a cutoff of 1060 HU for maximum CT attenuation, the AUC was 0.976, sensitivity was 95%, and specificity was 96%. The mean attenuation intraclass correlation coefficient was 0.987 for enostoses and 0.81 for metastases. The maximum attenuation intraclass correlation coefficient was 0.814 for enostoses and 0.980 for metastases.
CONCLUSIONS: CT attenuation measurements can be used to distinguish untreated osteoblastic metastases from enostoses. A mean attenuation of 885 HU and a maximum attenuation of 1060 HU provide reliable thresholds below which a metastatic lesion is the favored diagnosis.
摘要：
目的：这项研究的目的是确定CT衰减阈值是否可用于区分未经治疗的成骨细胞转移和成骨结瘤。
方法：研究组包括62例患者，在CT上发现279例硬化骨病变（37例患者中有126例结骨，25例患者中有153例转移）。通过组织学或临床和影像学随访评估硬化性病变的原因。没有患者接受过转移的预先治疗。以Hounsfield单位测量平均和最大衰减。进行ROC分析以确定灵敏度，特异性，AUC,95%CIs，和CT衰减的截止值，以区分转移瘤和结瘤瘤。使用具有95%CI的组内相关系数评估互读再现性。
结果：烯酮的平均和最大CT衰减值分别为1190±239HU和1323±234HU，分别,骨转移瘤为654±176HU和787±194HU，分别。使用885HU的截止值作为平均衰减，AUC为0.982，灵敏度为95%，特异性为96%。使用1060HU的截止值作为最大CT衰减，AUC为0.976，灵敏度为95%，特异性为96%。烯酮的平均衰减组内相关系数为0.987，转移为0.81。烯酮的最大衰减组内相关系数为0.814，转移灶的最大衰减组内相关系数为0.980。
结论：CT衰减测量可用于区分未治疗的成骨细胞转移和成骨结瘤。885HU的平均衰减和1060HU的最大衰减提供了可靠的阈值，在该阈值以下，转移性病变是有利的诊断。
方法：研究组包括62例患者，在CT上发现279例硬化骨病变（37例患者中有126例结骨，25例患者中有153例转移）。通过组织学或临床和影像学随访评估硬化性病变的原因。没有患者接受过转移的预先治疗。以Hounsfield单位测量平均和最大衰减。进行ROC分析以确定灵敏度，特异性，AUC,95%CIs，和CT衰减的截止值，以区分转移瘤和结瘤瘤。使用具有95%CI的组内相关系数评估互读再现性。
结果：烯酮的平均和最大CT衰减值分别为1190±239HU和1323±234HU，分别,骨转移瘤为654±176HU和787±194HU，分别。使用885HU的截止值作为平均衰减，AUC为0.982，灵敏度为95%，特异性为96%。使用1060HU的截止值作为最大CT衰减，AUC为0.976，灵敏度为95%，特异性为96%。烯酮的平均衰减组内相关系数为0.987，转移为0.81。烯酮的最大衰减组内相关系数为0.814，转移灶的最大衰减组内相关系数为0.980。
结论：CT衰减测量可用于区分未治疗的成骨细胞转移和成骨结瘤。885HU的平均衰减和1060HU的最大衰减提供了可靠的阈值，在该阈值以下，转移性病变是有利的诊断。
