Article Figures & Data
Figures
- Fig. 1.
Craniofacial bone marrow segmentation. The bone marrow of a 24.5-year-old man with sickle cell disease (outlined in blue) has been segmented on the axial proton attenuation–weighted images with 3-mm thickness. Approximately 75 images were segmented per subject. After segmentation was complete, 3D Slicer was used to layer the images over each other to create a 3D model. Shown below are sagittal (bottom left) and coronal (bottom right) images cut through the fully segmented cranial bone marrow after 3D reconstruction. After segmentation, the image data were analyzed in Mathcad for T1, T2, and secular-T2 relaxation times and volume.
- Fig. 2.
T1 and T2 histograms for craniofacial bone marrow. Mathcad-generated group histograms of craniofacial bone marrow T1 and T2 values for control subjects and subjects with SCD. Each vertical bar represents averaged relaxation times after segmentation. For T1, intersubject variability was high between groups, which prompted a more individualized analysis and breakdown of T1 peaks (Fig 3). For T2, the intersubject variability was minimal between the groups with more pronounced differences in relaxation times between subjects with SCD and control subjects.
- Fig. 3.
Individual histogram analysis of T1 peaks in a 32-year-old man without SCD (A) and a 42-year-old woman with SCD (B). A bimodal T1 peak distribution is seen in both cases, shown by dotted yellow and blue lines. Note the increased T1 shift seen in the subject with SCD and a reversal of T1 peak amplitudes compared with the control subject.
- Fig. 4.
Comparison of peak 1 (A) and peak 2 (B) components of T1 histogram. A, Significant increase of T1 peak 1 relaxation time is noted in patients with SCD, whereas there is no significant difference in peak 2 (B). Both T1 peaks in subjects with SCD showed an age-associated decline (P = .003 and P = .02 for T1 peak 1 and T1 peak 2, respectively).
- Fig. 5.
Comparison of T2 and secular-T2 peak relaxation times. Both T2 (A) and secular-T2 peak (B) relaxation time graphs show significant shortening of relaxation times in patients with SCD compared with control subjects. Age was not associated with either T2 or secular-T2 changes in subjects with SCD.
- Fig. 6.
T2 association with the number of transfusions received. Shortening of T2 times is seen as the number of packed red blood cell (PRBC) transfusions increase.
- Fig. 7.
Comparison of craniofacial bone marrow volume (BMV)/intracranial volume (ICV). Significant increase in craniofacial bone marrow volume is noted in patients with SCD compared with control subjects.
Tables
Parameter Mixed TSE Geometry Imaging plane Axial Acquisition matrix 256 × 192 Voxel dimensions (mm) 0.94 × 0.94 × 3.00 Intersection gap Null (2 packages) PE sampling (%) 75 FOV (FE) × FOV (PE) × mm2 240 × 180 Number of sections 80 Contrast Effective echo time (ms) TE1eff 7.142 TE2eff 100 TR (ms) 14,882.18 TI (ms) 1 700 2 7,441 Echo-train length 18 (9 per echo) Phase-encoding orders Centric 1st echo and linear 2nd echo Fat suppression No Acquisition Averages NEX = 1 SAR (W/kg) 2.7 Scan time (minutes) 9:05 Note:—SAR indicates specific absorption rate; FE, frequency encoding; PE, phase encoding.
Control (n = 17) SCD (n = 14) Analysis P Mean SD Mean SD T1 Peak 1 504.6 122.1 691.2 103.8 <.0001 T1 Peak 2 715.6 115.5 775.0 72.8 .09 T2 119.7 12.7 77.6 21.4 <.0001 Secular-T2 131.7 17.4 81.9 23.8 <.0001 BM volume 429.7 80.7 500.7 69.8 .01 BM/ICV 0.33 0.07 0.38 0.06 .02 Note:—BM indicates bone marrow; ICV, intracranial volume.
↵a Not shown: there was an age association between T1 peak 1 times (P = .003) and T1 peak 2 times (P = .02).
