Combined intravoxel incoherent motion diffusion-weighted MR imaging and magnetic resonance spectroscopy in differentiation between osteoporotic and metastatic vertebral compression fractures

Abstract

Purpose Our purpose was to combine intravoxel incoherent motion diffusion-weighted MR imaging (IVIM-DWI) and magnetic resonance spectroscopy (MRS) to differentiate osteoporotic fractures from osteolytic metastatic vertebral compression fractures (VCFs). Methods A total of 70 patients with VCFs were included and divided into two groups, according to their causes of fractures based on pathological findings or clinical follow-up. All patients underwent conventional sagittal T1WI, T2WI, STIR, IVIM-DWI, and single-voxel MRS. The diffusion coefficient (D), pseudo diffusion (D*), and perfusion fraction (f) parameters from IVIM-DWI and the lipid water ratio (LWR) and fat fraction (FF) parameters from MRS were obtained and compared among groups. Furthermore, the diagnostic performance of MRS, IVIM-DWI, and IVIM-DWI combined with MRS for differentiation between osteoporotic and osteolytic metastatic VCFs was assessed by using receiver operating characteristic (ROC) curve analysis. Results Compared with the osteoporotic group, the metastatic group had significantly lower values for f, D, and FF, but higher D * (all P < 0.05). The area under the receiver operating characteristic (ROC) curve of MRS, IVIM-DWI, and IVIM-DWI combined with MRS were 0.73, 0.88, and 0.94, respectively. Among these, the IVIM-DWI combined with MRS showed the highest sensitivity, specificity, and accuracy, which are 90.63% (29/32), 97.37 % (37/38), and 94.29% (66/70), respectively. Conclusions IVIM-DWI combined with MRS can be more accurate and efficient for differentiation between osteoporotic and osteolytic metastatic VCFs than single MRS or IVIM-DWI.

Different Attenuation Models of Diffusion‐Weighted MR Imaging for the Differentiation of Benign and Malignant Musculoskeletal Tumors

“…The number of studies that assess the value of IVIM and DK in the differentiation of benign and malignant musculoskeletal tumors is fewer than those using the monoexponential diffusion model. [8][9][10][11][12][13][14][15][16][17][18][19] In this study, we found a statistically significant difference in D ivim-min between benign and malignant tumors in the "all tumors" group and in the nonmyxoid and nonchondroid tumor subgroup. The best cut-off value was 0.999 Â 10 À3 mm 2 s À1 for D ivim-min in both tumor groups.…”

“…In addition, the interobserver and intraobserver agreement was moderate for f min and poor for D* min . In the current literature, most of the studies, which assess the value of IVIM-derived parameters in musculoskeletal tumors are performed at 3 T, and ICCs varied between 0.83-0.947 for D* and 0.81-0.957 for f. [8][9][10]13,18 In one exceptional study which was performed at 3 T ICCs were 0.386-0.764 for D* and 0.588-0.841 for f. 11 In one study performed at 1.5 T with 12 b values (9 b values < 200 mm 2 s À1 ), ICCs were 0.786 for D* and 0.701 for f, 8 which were lower than many of the studies conducted at 3 T, yet still higher than our findings. The discrepancy with our results might be related to lower SNR due to our lower field strength, lower number of b values (11 b values and 8 of them < 200 mm 2 s À1 ), and possibly lower averages used in order to keep the overall examination time as short as possible (b values/averages: 0/2, 10/2, 20/2, 40/2, 50/2, 75/2, 100/2, 150/2, 300/2, 500/2, and 800/8).…”

“…Third, an overlap in ADC values of benign and malignant tumors can be seen due to fibrous stroma or large extracellular matrix. 5,7,22,23 In contrast to the previously published studies which assessed the per group performance of the minimum or mean values of m-DWI-, 7,27-30,37 IVIM-, [8][9][10][11][12][13][14][15] and DKderived [16][17][18][19] parameters, our study additionally used a volumetric analysis of the entire tumor volume, including cystic/necrotic or hemorrhagic areas. Given that the absence of a standard ROI placement technique results in difficulties in comparing similar studies, VOI measurements could be useful and less susceptible to interobserver variability when compared to ROI measurement techniques.…”

“…4 DK-derived mean diffusivity (D K ) represents the corrected diffusion coefficient according to such non-Gaussian behavior, and the mean kurtosis (K) value quantifies the deviation of molecular diffusion from the Gaussian distribution. 4 Although a variety of functional imaging techniques, including m-DWI, IVIM and DK, have been used [5][6][7][8][9][10][11][12][13][14][15][16][17][18][19] in differentiating benign and malignant musculoskeletal tumors, all three techniques have not been studied in the same patient population.…”

Association between bone turnover markers, BMD and height loss of cemented vertebrae after percutaneous vertebroplasty in patients with osteoporotic vertebral compression fractures

“…Moreover, the follow-up period is relatively short, while a longer follow-up time is necessary to determine the risk factors for height loss of treated vertebral bodies after PVP. Some new technologies, such as combined intravoxel incoherent motion diffusion-weighted MR imaging and magnetic resonance spectroscopy, have not been employed to differentiate osteoporotic fractures from osteolytic metastatic vertebral compression fractures in this study [ 41 ]. Finally, every patient’s daily activity is not controlled entirely in follow-up period after PVP.…”