Research archive · 2019 journal issue · Online publication 2018
Preprocessing choices are part of the measurement—not just a preparation step.
The study “Beware detrending” by Woletz and colleagues, co-authored by Martin Tik, compared preprocessing strategies for low-frequency fluctuation analysis in resting-state fMRI. It examined how analysis choices influence the resulting measures, including ALFF and fractional ALFF (fALFF).
ALFF describes the amplitude of low-frequency signal fluctuations; fALFF expresses the low-frequency contribution relative to a broader frequency range. These measures characterise fluctuation amplitude. They are distinct from functional connectivity, which describes relationships between signals from different regions.
Connection to current research
Careful handling of motion, physiological effects and preprocessing remains important when using fMRI to characterise networks or interpret stimulation-related responses. The original paper provides the tested procedures and their limitations.
Woletz et al. Beware detrending: Optimal preprocessing pipeline for low-frequency fluctuation analysis. Human Brain Mapping, 40, 1571–1582 (2019). Abstract and full author list ↗
Resting-state fMRI → · Different kinds of brain connectivity →