Differentiation of multiple myeloma and metastases with apparent diffusion coefficient map histogram analysis

Abstract

OBJECTIVE: Multiple myeloma and metastasis are common malignant bone marrow lesions. It may be difficult to distinguish from each other due to similar radiological findings. This study aimed to determine the usefulness of histogram analysis with diffusion-weighted imaging (DWI) in the differentiation of multiple myeloma and metastasis. METHODS: Twenty patients with multiple myeloma and 20 patients with metastasis who underwent 3T magnetic resonance (MR) imaging with DWI (b=0, 1000 s/mm(2)) were enrolled. All patients had multiple enhancing nodular bone lesions on contrast-enhanced musculoskeletal MR imaging. Histogram analysis was performed from these lesions on the apparent diffusion coefficient (ADC) map. The mean, minimum, median, maximum, standard deviation of the histogram, variance, entropy, uniformity, skewness, kurtosis, size %lower, size %upper, and size %mean values were measured. Results of both groups were compared. RESULTS: The mean, minimum, median, maximum, standard deviation, and variance values were found to be significantly lower in multiple myeloma than metastasis (p<0.001). When ROC analysis was performed for mean value, the area under the curve=1.000 and when threshold value was selected as 766.076, two groups could be differentiated with 100.0% sensitivity and 100.0% specificity. CONCLUSION: ADC histogram analysis can be considered as a method to be used in the differentiation of metastases and multiple myeloma.