Sources of error in the quantitative analysis of MRI scans

The increasing use of quantitative analysis of MRI scans in the literature has produced a need to identify potential sources of bias in such analysis procedures. Six sources of bias are demonstrated in this paper. These include bias attributable to partial volume effects, head tilt, plane of view, use of noncontiguous slices, contrast/intensity manipulations, and magnetic inhomogeneities. The magnitude of bias for each source varied according to whether a hemisphere or regions within a hemisphere were measured, with regional effects typically exceeding hemisphere effects.