PubMed Health⌕ Search

Biomedical subjects

P F Gilbert

Publications and source records attributed to P F Gilbert.

12 recordsLinked to original sources
PubMed2001

An outline of brain function.

An outline of how the brain may compute is proposed. In the cerebral cortex memories are stored through long-term potentiation at synapses from layer 1 cortical inputs (representing contexts) on layer 2/3 pyramidal cells linked with the thalamus in a cortico-thalamic (CT) unit. The signals which are memorized are the layer 3 inputs from the thalamus or other cortical areas. Signals are memorized (and later recalled) at the gamma frequency. A conscious thought comprises the outputs of layer 5 cells in CT units in different cortical regions firing in synchrony through the contribution of oscillatory thalamic and cortical inputs. This cortical output influences sub-cortical areas to cause or participate in a movement. Cerebral cortical outputs may be stored in the cerebellum and generated later in a particular context by the basal ganglia and cerebellum. Thus the brain may either generate 'conscious' outputs using the cerebral cortex or 'automatic' outputs using the basal ganglia and cerebellum. When contexts are recognized by the basal ganglia it permits outputs stored in the cerebellum to commence and in this way the basal ganglia can control complex sequences of outputs or movements. Working memory involves the prefrontal cortex using similarly the basal ganglia and cerebellum. The hippocampus has a role in the storage and recall of cortical outputs by providing unique layer 1 contexts to all the CT loops in different cortical areas in a conscious thought. With further recall of the thought new layer 1 contexts may become associated with the CT loops enabling recall without the hippocampal input.

Animals↗
PubMed1978

Convergence of cerebral inputs onto dentate neurons in monkey.

The patterns of convergence of inputs from different areas of the cerebral cortex and the peripheral nerves onto single dentate neurons was studied in cebus monkeys. Dentate neurons receive their strongest and most numerous inputs from the premotor and supplementary motor regions of area 6. The sensorimotor and frontal cortices have weaker projections to the dentate nucleus, while peripheral nerves and many other association cortical areas were found to be effective in influencing cells of the lateral cerebellum. Dentate cells that respond to stimulation of hindlimb regions of the sensorimotor cortex tend to receive their principal input from the supplementary motor area and medial premotor regions, while neurons responding to forelimb sensorimotor cortex tend to receive lateral premotor inputs. In addition there is a topographical organization within the ventral pole of dentate with the hindlimb represented in the anterior regions and the forelimb in the posterior regions. These results are compared with those of similar studies of interpositus and dentate neurons in cat and monkey. The differences between the afferent inputs to dentate and interpositus are consistent with the suggestion that the lateral cerebellum is involved in programming movement parameters before movement initiation while the intermediate zone is involved in up-dating the evolving movement.

Animals↗
PubMed1977

Purkinje cell activity during motor learning.

Monkeys were trained to grasp a handle and move it in a horizontal arc to a central position by flexing or extending the wrist. A torque motor applied forces to the handle that switched at random intervals to alternately load flexor and extensor muscles. At each load switch, the handle was displaced transiently from the central position, and then moved back by the monkeys and held there steadily again. Recordings were made from cerebellar Purkinje cells (P-cells) whose simple spike (SS) activity was related to the task. The magnitude of one of the oppositely directed loads was then altered and the monkeys took about 12-100 trials with the novel load before performing as regularly as previously. During this period with one known and one novel load, some P-cells underwent increases in complex spike (CS) frequency at specific times after the load switch. This increased CS frequency would last for a similar number of trials as that taken by the monkey to adapt to the novel load before decreasing to near its previous level. Associated with the increased CS frequency ther were decreases in SS frequency that persisted after the CS frequency had decreased to near its previous level. These results are consistent with theoretical proposals that motor learning takes place in the cerebellum through changes in the strength of transmission of parallel fiber synapses on P-cells caused by the climbing fiber input. These results further suggest that climbing fiber firing causes a decrease in the strength of parallel fiber synapses.

Action Potentials↗
PubMed1977

Integration of cerebral and peripheral inputs by interpositus neurons in monkey.

The patterns of convergence of cerebral and peripheral nerve inputs onto interpositus neurons were studied in cebus monkeys. The strongest inputs to interpositus neurons are from motor and somatosensory cortex, with weaker inputs from peripheral nerves and cerebral area 6. The neurons in the anterior portion of interpositus receive cerebral and peripheral inputs primarily representing the hindlimb, while inputs to neurons in the posterior division represent forelimb or mixed forelimb and hindlimb. The hindlimb neurons integrate signals principally from motor cortex, somatosensory cortex, nerves, supplementary motor and medial pre-motor areas, while forelimb neurons receive inputs from motor, somatosensory, lateral premotor cortical areas and nerves. The results from this study are compared with those from studies of interpositus and dentate neurons in cat and monkey in order to determine the role of n. interpositus in movement. It is suggested that the inputs integrated by interpositus neurons are consistent with a role in up-dating skilled movements.

Action Potentials↗