Determination of bone marrow fat fraction with 1.5 T MRI and the role of fat fraction in discrimination of bone marrow malignities in children: A retrospective study

One of the most valuable methods for noninvasive evaluation of bone marrow is MRI. Although conventional sequences allow for qualitative evaluation, qualitative evaluation can be quite difficult, especially in the pediatric patient population, due to the variability of red bone marrow with age. Newly developed MRI sequences such as MRI - FF and Dixon method can provide quantitative data about the structure of the bone marrow, allowing for more accurate evaluation of the bone marrow. In our study, it was aimed to determine the contribution of the data obtained from conventional sequence, Dixon method and MRI-FF to the diagnosis in the presence of diffuse bone marrow infiltration in the pediatric population. 79 pediatric patients who underwent lumbar and pelvic imaging were evaluated retrospectively. These patients were divided into 3 groups according to their bone marrow findings. The first group included 10 patients with pathologically proven malignant infiltration of the bone marrow, the second group included 32 patients without bone marrow abnormalities on MRI, and the third group included 6 patients with benign pathologies in the bone marrow. In addition, 63 patients without bone marrow pathology were divided into 4 groups according to their age. ROI was placed in L4 and L5 vertebrae, bilateral iliac wings and bilateral proximal metaphyses of the femur in T1W, out of phase/in phase and MRI-FF images and measurements were made. Measurements (T1 SI, SIO and FF) from patients in the malignant, benign and control groups were compared. When the T1 SI values obtained from six regions are examined; there was a significant difference between the malignant and normal groups in the other four localizations, excluding the iliac wings. There was a significant difference between the benign and normal groups in the other four localizations, excluding the vertebrae. There was no significant difference between the malignant group and the benign group in any localization. When the SIO was examined, a significant difference was found between the malignant and normal groups in all regions. When the malignant and benign groups were compared, a significant difference was found only in the proximal parts of the femur. A significant difference was found between the benign and normal groups in other localizations, except for the femoral regions. When the FF values were examined, a significant difference was found between the malignant and benign groups and the normal group in all regions. There was no significant difference between the malignant group and the benign group. The cut-off values found for each site in malignant infiltration of the bone marrow ranged from 8.00-14.00 for FF and 0.76-0.87 for SIO. For T1 SI, a significant cut-off value could not be determined in the measurements made from both iliac wings. The cut-off values for L4-L5 vertebrae and both femurs varied between 88.00-159.00. When the images of the patients in four groups were examined to determine the change in FF value with age, no significant difference was found between the FF values of any region. In conclusion, MRI-FF can be considered as a method with high sensitivity and specificity in differentiating benign and malignant pathologies that affect the bone marrow diffusely from normal bone marrow; however, it was not found to be useful in differentiating malignant involvement from bone marrow reconversion. Unlike the other two parameters evaluated in our study, SIO was found to be effective in differentiating malignant involvement from bone marrow reconversion even if only in the proximal femoral metaphyses. It is also useful in differentiating malignant and benign pathologies. Diffuse bone marrow abnormalities are usually first noticed on T1W images; however, the sensitivity and specificity of this sequence is low compared to other techniques. For this reason, quantitative evaluation with MRI-FF or SIO may be more useful methods. Keywords: Bone marrow, Magnetic Resonance Imaging, Fat fraction