Changes of the apparent diffusion coefficient in brain diffusion-weighted images due to subject positioning: A simulation study

Background and purpose: The purpose of this work was to investigate whether the cortical apparent diffusion coefficient (ADC) values derived from brain diffusion-weighted images vary with changes in the position of the subject against a static magnetic field. Materials and methods: To focus on the variations in ADC due to the change of subject positioning, a simulation was performed using a digital brain phantom. The magnetic field inhomogeneities in the digital phantom were calculated for each subject position while changing the angle between the direction of the static field and the head of the digital phantom. The angle was changed from 0 to 40 degrees at 10-degree intervals. For each angle, the diffusion-weighted images were simulated based on magnetic resonance physics in which the magnetic field inhomogeneity was taken into account. The relative differences of average ADC values between the tilt angles were calculated to evaluate the variations in ADC. The Wilcoxon rank-sum test was used for comparisons of ADC values between the tilt angles for each cortical region. Results: In the cortical regions distorted by magnetic field inhomogeneities, the average ADC values differed significantly according to the position of the subject (P<. 0.05). The range of the relative differences in average ADC values in relation to the differences in subject positioning was approximately 1% to 12%. Conclusion: Our results suggest that subject positioning against a static field is one of the factors affecting the accuracy of cortical ADC measurements derived from brain diffusion-weighted images.