Peak-Interval Timing in Humans Activates Frontal-Striatal Loops S.C. Hinton*, W.H. Meek*, J.R. MaeFallt *Duke University, Department of Psychology: Experimental, Durham, NC, USA ?Duke University Medical Center, Department of Radiology, Durham, NC, USA Introduction Research on human timing to date has largely been confined to behavioral studies of normal human beings (1) although attempts have been made to study timing in patient populations (2). Converging evidence from studies in animals and humans suggests that interval timing depends on activation of neural circuits through the frontal cortex, striatum, and thalamus that are described as frontal-striatal loops (3). A successful information-processing model of time perception has guided this research for many years (4,5), and this model has lately been applied to human timing data as well (6). Still, little is yet known anatomically about how the human brain processes time, and the current study seeks to address this issue using functional magnetic resonance imaging (fMRI). Subjects Four normal male right-handed adults between the ages of 18 and 60. Methods A single-shot echo-planar pulse sequence (40 x 20 cm FOV, 60 ~ flip angle, 1000 ms TR, 50 ms TE, 128 x 64 pixel matrix, 250 KHz bandwidth) was used to obtain functional MR images on a 1.5T scanner (Signa, GE Medical Systems with ANMR InstaScan accessory). Subjects' heads were aligned with the cantho-meatal line vertical. Four contiguous 7 mm axial slices were sampled each second, and the most inferior slice lay approximately along the bicommissural axis. Testing sessions were composed of parametric variations of a behavioral task known as the peak-interval timing procedure (7), and functional images were collected of the resulting brain activity while subjects performed the task. They were initially trained with five fixed-time signal trials of 11 sec, and functional images were then acquired continuously while the subject was tested for eight 20-sec probe trials which were each preceded by a 10-sec intertrial interval (ITI). The signal to be timed was the presence (foreground) or absence (background) of either an auditory stimulus (440 Hz tone presented through headphones) or a visual stimulus (32 Hz flashing red light-emitting diodes in goggles mounted on the head coil). In the foreground condition, the stimulus was on during the probe trial and off during the ITI, whereas the reverse was true in the background condition. During control trials, the subject passively attended to the stimulus. In experimental trials, the subjects' task was to make two timecontrolled responses by squeezing a sphygmomanometer bulb such that the first response came just before and the second response came just after their subjective estimate of the 11-sec criterion time. A timing control condition had subjects time the signal without making a motor response. During a motor control condition, the subject was instructed by the experimenter when to make two responses during each trial that were not controlled by time. Results Correlation coefficient maps were produced for each condition to compare changes in brain activation during probe trials relative to the ITI. Subtraction of correlation maps obtained under different conditions allowed isolation of pure interval timing from motorand sensory-specific activation. Timing-related activation was found in the striatum, thalamus, and frontal cortex. Conclusion These data show involvement of frontal-striatal loops in timing of short intervals by humans and represent the first reported evidence for task-specific involvement of the basal ganglia in a cognitive task requiring temporal processing.
Read full abstract