医学
核医学
骨髓
病变
T2加权
神经组阅片室
磁共振成像
病理
放射科
神经学
精神科
作者
Sekyoung Park,Jin Do Huh
标识
DOI:10.1007/s00330-022-08965-3
摘要
To differentiate bone metastases (BMs) from benign red marrow depositions (BRMs) of the spine using quantitative parameters derived from fat-suppressed T2-weighted imaging (T2 FS) and fat fraction (FF) map METHODS: One hundred eleven lesions, divided into 62 BMs and 49 BRMs according to MR images and either bone scan or PET-CT, were assessed with T2 FS and FF map. Two radiologists independently measured quantitative parameters from the ROIs in the lesions, including fat-suppressed (FS) T2 ratio (ratio of lesion FS T2 signal intensity [SI] to normal marrow FS T2 SI), FF, and FF ratio (ratio of lesion FF to normal marrow FF). The mean values of these parameters were compared between the two groups. To evaluate the diagnostic utilities of individual (FS T2 ratio, FF, and FF ratio) and combined parameters, ROC curves were analyzed. For the ROC curves among the individual parameters and their combinations, AUCs were compared.The FS T2 ratio of BMs was significantly higher than that of BRMs (2.638 vs. 1.155 [p < 0.001]). The FF and FF ratio of BMs were significantly lower than those of BRMs (FF, 3.554% vs. 20.038% [p < 0.001]; FF ratio, 0.072 vs. 0.364 [p < 0.001]). The ROC AUCs of individual and combined parameters ranged from 0.941 to 0.980. The AUCs of all individual parameters and their combinations did not demonstrate statistically significant differences.The FS T2 ratio, FF, and FF ratio can be useful in differentiating BMs from BRMs with or without any combination of the parameters.• Quantitative parameters derived from fat-suppressed T2-weighted imaging and fat fraction map could be used to differentiate bone metastases from benign red marrow depositions with or without any combination of the parameters. • Quantitative parameters of fat-suppressed T2-weighted imaging provide diagnostic performance similar to those of fat fraction map in differentiating bone metastases from benign red marrow depositions.
科研通智能强力驱动
Strongly Powered by AbleSci AI