Magnetic resonance imaging based bone marrow segmentation for quantitative calculation of pure red marrow metabolism using 2-deoxy-2-[F-18]fluoro-D-glucose-positron emission tomography: A novel application with significant implications for combined structure-function approach

Aims: The aim of this study was to introduce a new concept for accurate measurement of the global metabolic activity of the red marrow by combining segmented volumetric data from structural imaging techniques such as magnetic resonance imaging (MRI) and quantitative metabolic information provided by functional modalities such as positron emission tomography (PET). Materials and Methods: Imaging studies from five subjects who had undergone both MRI and 2-deoxy-2-[F-18]fluoro-D-glucose(FDG)-PET were selected for this analysis to test the feasibility of this approach. In none of the subjects, there were any marrow abnormalities as determined either by the MRI or by FDG-PET studies. The mean blood glucose level was 96 +/- 25 mg/dl. The first step was to calculate vertebral volume at L1, L3, and L5 from the available MRI studies. The red and yellow marrows were then segmented within the lumbar vertebrae using the 3DVIEWNIX software system from which the respective volumes were also calculated for each. This also allowed calculating the bone volume in each of the vertebral bodies examined. By employing the standard techniques, the mean of the maximum standardized uptake values (mean SUVmax) for the bone marrow were calculated in L1, L3 and L5 of the lumbar spine, and then global red marrow activity was calculated using the following approach: (1) Whole vertebral metabolic activity (WVMA)=vertebral volume x emean SUVmax of the entire marrow, (2) whole vertebral metabolic activity for yellow marrow (WVMAYM)=eyellow marrow volume x mean SUVmax of fat (obtained from measurements of subcutaneous fat), (3) whole vertebral metabolic activity for red marrow (WVMARM)=WVMA - WVMAYM; and finally, (4) SUVmax for pure red marrow=whole vertebral metabolic activity for red marrow (WVMARM)/red marrow volume (obtained from the segmentation data). Results: The mean volume of the lumbar vertebral body was 15.6 +/- 1.4 cm(3), the bone marrow mean SUVmax was 1.5 +/- 0.3, and the MVP for the lumbar vertebral body was 23.4 +/- 5.9. The mean volume of the yellow marrow in the lumbar vertebral body was 7.7 +/- 1.1 cm(3), the yellow marrow mean SUVmax was estimated to be 0.38 +/- 0.1 and the MVP for the yellow marrow in the lumbar vertebral body was 2.9 +/- 0.9. The mean volume of the red marrow in lumbar vertebral body was 7.9 +/- 1.1 cm(3), the red marrow mean SUVmax was estimated to be 2.6 +/- 0.6, and the MVP for the red marrow in the lumbar vertebral body was 20.5 +/- 5.9. Conclusion: Estimation of the individual component of the bone marrow is plausible using medical image segmentation with combined structure-function approach. This can have potential research and clinical applications concerning the study of global metabolic activity of the individual component and diagnosis of benign and malignant bone marrow disorders.