Fig. 2.

75-year-old woman with a benign vertebral compression fracture (VCF) (a–j). Sagittal thoracic spine MR images generated from a single FSE T2-weighted Dixon sequence include in-phase (a and f), out-of-phase (b and g), fat-only (c and h), water-only images (d and i), and fat fraction (FF) maps (e and j). Region of interest (ROI)1 (a–e) was drawn on the entire vertebra at the level of the VCF on (a) in-phase image and copy-pasted onto the (b) out-of-phase, (c) fat-only, (d) water-only images, and (e) fat fraction maps. ROI2 (f–j) was drawn on the area with high signal intensity on the (i) water-only image, and copy-pasted onto the (f) in-phase, (g) out-of-phase, (h) fat-only images, and (j) fat fraction maps. Control ROI (circle) placed on an adjacent healthy vertebra and is depicted on all images. In this benign VCF, signal drop, FF calculated from water-only and fat-only images, and FF from maps were ROI1: 50.4%/ROI2: 47.6%; ROI1: 51.4%/ROI2: 36.2%; and ROI1: 39.8%/ROI2: 31.3%, respectively