Background: As a potential brain imaging biomarker, amplitude of low frequency fluctuation
(ALFF) has been used as a feature to distinguish patients with Alzheimer’s disease (AD)
and amnestic mild cognitive impairment (aMCI) from normal controls (NC). However, it remains
unclear whether the frequency-dependent pattern of ALFF alterations can effectively distinguish
the different phases of the disease.
Methods: In the present study, 52 AD and 50 aMCI patients were enrolled together with 43 NC in
total. The ALFF values were calculated in the following three frequency bands: classical
(0.01-0.08 Hz), slow-4 (0.027-0.073 Hz) and slow-5 (0.01-0.027 Hz) for the three different groups.
Subsequently, the local functional abnormalities were employed as features to examine the effect
of classification among AD, aMCI and NC using a support vector machine (SVM).
Results: We found that the among-group differences of ALFF in the different frequency bands
were mainly located in the left hippocampus (HP), right HP, bilateral posterior cingulate cortex
(PCC) and bilateral precuneus (PCu), left angular gyrus (AG) and left medial prefrontal cortex
(mPFC). When the local functional abnormalities were employed as features, we identified that the
ALFF in the slow-5 frequency band showed the highest accuracy to distinguish among the three
Conclusion: These findings may deepen our understanding of the pathogenesis of AD and suggest
that slow-5 frequency band may be helpful to explore the pathogenesis and distinguish the phases
of this disease.