Purpose: Diffusion-weighted imaging (DWI) has become an important component in modern stroke imaging. This MR technique detects diffusion abnormalities, which can be quantified by computing apparent diffusion coefficient (ADC) maps. ADC values are typically calculated from a set of MR images obtained with varying degrees of diffusion weighting (b-values) using nonlinear regression. However, there is no agreement concerning the number of images needed for ADC calculation. The aim of our study was to determine how many b-values are necessary to reliably calculate ADC maps.
Materials and methods: In 100 consecutive patients with clinical signs of acute ischemic stroke, 6 identically oriented and centered diffusion data sets with different b-values were acquired. ROI analysis was performed for DWI-positive lesions, normal-appearing gray and white matter, CSF, and background noise. ADC values for each ROI were calculated using a nonlinear regression model. Additionally, the CNR and SNR were calculated for each ROI.
Results: Acquisition time was 0:39 min for 2 b-values and up to 2:49 min for a sequence with 7 b-values. The mean ADC (× 10(-3) mm2/s) for ischemic lesions was 58.29, 58.47, 57.83, 57.81, 57.58 and 54.51 using 2, 3, 4, 5, 6, and 7 b-values. Ischemic lesions had significantly different mean ADC values only for high b-values (b = 2000 s/mm2).
Conclusion: ADC values can be reliably calculated using 2 b-values. Radiologists may use the more time-efficient 2-point method for reliably estimating ADC values and detecting ischemic lesions in the daily clinical routine.
© Georg Thieme Verlag KG Stuttgart · New York.