Fractional anisotropy--threshold dependence in tract-based diffusion tensor analysis: evaluation of the uncinate fasciculus in Alzheimer disease.

BACKGROUND AND PURPOSE Tract-based analysis can be used to investigate required tracts extracted from other fiber tracts. However, the fractional anisotropy (FA) threshold influences tractography analysis. The current study evaluated the influence of the FA threshold in measuring diffusion tensor parameters for tract-based analysis of the uncinate fasciculus in subjects with Alzheimer disease (AD). MATERIALS AND METHODS Subjects included 30 patients with AD and 10 healthy controls. We acquired tractographies of the uncinate fasciculus by using different FA thresholds. We measured mean FA and the apparent diffusion coefficient (ADC) along the uncinate fasciculus for different FA thresholds and evaluated the correlation between diffusion tensor parameters (FA, ADC) and the Mini-Mental State Examination (MMSE) scores. RESULTS The uncinate fasciculus showed lower mean FA and higher mean ADC values in cases with more severe AD. A higher FA threshold led to a lower mean ADC value and a higher mean FA value along the uncinate fasciculus, whereas the relative order of measured values according to the severity of AD was not influenced by the FA threshold. An FA threshold of 0.2 showed higher correlation between mean ADC values and MMSE scores. FA thresholds of 0.15 and 0.20 showed higher correlation between mean FA values and MMSE scores. CONCLUSIONS Appropriate selection of the FA threshold leads to higher correlation between diffusion tensor parameters and the severity of AD. For tract-based analysis of degenerative diseases such as AD, appropriate selection of the FA threshold for tractography is important.