Parametric analysis of functional neuroimages: application to a variable-rate motor task.

Positron emission tomography (PET) has proven to be a powerful tool in identifying the functional neuroanatomy underlying cognitive and sensorimotor processing. In this paper, we present a method for mathematically modeling the changes in regional cerebral blood flow (rCBF) as a function of experimental parameters using step and linear functions. PET was used to measure rCBF in six subjects who tracked a target moving with constant amplitude across a computer screen at four different frequencies. Each subject tracked the target by flexing and extending the wrist. Two scans were performed at each frequency. The data for each subject were normalized by the mean blood flow in each scan and scaled to the mean blood flow at rest. Scaled rCBF was regressed onto movement frequency to identify voxels which had either a significant linear or step function response to the frequency of movement. A group analysis was also performed to identify significant functional changes common to all subjects. Significant rCBF increases in relation to movement frequency were found in the supplementary motor area, primary motor cortex, premotor cortex, thalamus, and cerebellum and localized using the Talairach atlas. Habituation of responses was not observed.