Functional MRI activation signal detection using the periodicity transform

Abstract

A key challenge in Functional Magnetic Resonance Imaging (fMRI) is the detection of activation areas in the brain. We introduce a new method of M R I signal detection, using an approach termed the Periodicity Transform. The technique is based on temporal data analysis. A search for periodicity is camed out in the tMRI time series data. The method is applicable to block design experiments. In the block paradigm, the stimulus period is known and it is possible to use this information for searching periodicities in the time series data. We present the results for the periodicity detection in the time series of the simulated phantom as well as clinical fMR1 data from the finger tapping experiment. No assumptions have been made about the amplitude and frequency of the activation signal. The algorithm extracts arbitrary harmonics at the periodicity defined by thc stimulus function.