The role of cerebral cortex in the generation of voluntary saccades: A positron emission tomographic study

The purpose of this study was to define the location and behavior of cerebral structures within the normal human brain that participate in the generation of voluntary saccadic eye movements. Changes in regional cerebral blood flow (rCBF) during task performance were assumed to reflect like changes in regional neuronal activity induced by the task. The locations of all rCBF changes were described in stereotaxic coordinates. Cerebral blood flow (CBF) was measured with positron emission tomography (PET) and bolus intravenous injection of H₂¹⁵O. The use of H₂₁⁵O with PET allowed six, seven-slice measurements of brain blood flow to be made in rapid sequency for each subject, without removing the subject from the tomograph between scans. Nine paid normal volunteers were studied. The paradigm included three saccadic eye-movement (SEM) conditions, one finger-movement condition and two control conditions (initial and final). The three SEM conditions allowed comparisons to be drawn between targeted versus untargeted SEMs, auditorily cued versus visually cued SEMs and stochastic versus rhythmic SEMs. All tasks were simple and deterministic in that each movement exactly mirrored the preceding movement: finger flexion then extension, saccade-left then saccade-right. Saccadic eye movements were associated with rCBF increases within the frontal eye fields, the supplementary motor area, and the cerebellum. Finger movements were associated with rCBF changes within the sensorimotor hand areas, the supplementary motor area, and the cerebellum. The frontal eye fields were discrete cortical regions consistently active during the generation of voluntary SEMs and uninfluenced by target presence, type of cue, or task complexity, indicating a predominantly motor function. The supplementary motor area (SMA) was consistently active during all motor tasks and was uninfluenced by the degree of task complexity or stochasticity. A role for SMA in establishing 'motor set' during both simple and complex motor tasks is suggested. An anterior-posterior somatotopy was found for SMA-eye (anterior) versus SMA-hand (posterior). Lateral occipital visual association cortex activation was present only during targeted saccadic conditions. Comparative analysis with prior rCBF measurements made during visual stimulation in the absence of saccades, indicated that the rCBF increase was due to target presence not oculomotor activity.