Quantifying bone marrow fat using standard T1-weighted magnetic resonance images in children with typical development and in children with cerebral palsy

Excess bone marrow adiposity may have a negative effect on bone growth and development. The aim of this study was to determine whether a procedure using standard T1-weighted magnetic resonance images provides an accurate estimate of bone marrow fat in children with typical development and in children with mild spastic cerebral palsy (CP; n = 15/group; 4–11 y). Magnetic resonance imaging was used to acquire T1-weighted images. It was also used to acquire fat and water images using an iterative decomposition of water and fat with echo asymmetry and least-squares estimation (IDEAL) technique. Bone marrow fat volume and fat fraction in the middle-third of the tibia were determined using the standard T1-weighted images (BMFV(T1) and BMFF(T1), respectively) and the fat and water images (BMFV(IDEAL) and BMFF(IDEAL), respectively). In both groups, BMFV(T1) was highly correlated with (both r > 0.99, p < 0.001) and not different from (both p > 0.05) BMFV(IDEAL). In both groups, BMFF(T1) was moderately correlated with (both r = 0.71, p < 0.01) and not different from (both p > 0.05) BMFF(IDEAL). There was no group difference in BMFV(T1) or BMFV(IDEAL) (both p > 0.05). BMFF(IDEAL) was higher in children with CP (p < 0.05), but there was no group difference in BMFF(T1) (p > 0.05). We conclude that a procedure using standard T1-weighted magnetic resonance images can produce estimates of bone marrow fat volume similar to estimates from the IDEAL technique in children. However, it is less sensitive to variation in the bone marrow fat fraction.