Collin Liu1, 2, Anitha Priya Krishnan3, Lirong Yan4, Jeffrey R. Alger4, John Ringman5, Danny JJ Wang1, 6
1Ahmanson-Lovelace Brain Mapping Center, UCLA , Los Angeles, CA, United States; 2Neurobehavior Unit, VA Greater LA Healthcare System, Los Angeles, CA, United States; 3Molecular Imaging Center, USC, Los Angeles, CA; 4Ahmanson-Lovelace Brain Mapping Center, UCLA, Los Angeles, CA, United States; 5Neurology, UCLA, Los Angeles, CA; 6Neurology, UCLA, Los Angeles , CA
Interpretation of biological signals is essential for diagnosing diseases. Pattern recognition and spectral analyses have commonly been used. More recently a non-linear time-series analysis called approximate entropy has been applied to EEG, ECG, and hormonal levels. Here we applied approximate entropy to resting state BOLD fMRI time-series, to characterize the complexity of the signal in normal aging and familial Alzheimer's disease. Similar calculation can be made between the time-series of a seed voxel and that of all other voxels to provide a measure of synchronicity. This is called cross-approximate entropy. These analyses might provide novel measures of functional connectivity, complementary to cross-correlation.