The adaptive abilities of the cerebral cortex begin at the cellular and subcellular level. We will examine how individual neurons are capable of information processing and memory.
Neurons display the following properties:
integration and processing of inputs
functional adaptations based on internal and external environmental conditions
Biological Logic Gates
Neurons have a integrative function similar to logic gates in computers, although immensely more complex. A neuron receives a range of input signals through its dendrites, integrates them, and then may or may not “fire”, depending on the type and frequency of the input signals. The firing of a neuron propogates signals to other neurons, and is generally an all or nothing process; the inputs must excite the neuron beyond a certain threshold for it to fire. The input signal depends on whether the connecting synapse is strong or weak, excitatory or inhibitory. A neuron with two inputs can act in several different modes, depending on the types and strengths of its inputs.
As a simplification, we will only look at logical operations with just two inputs, but the model generalizes to an arbitrary number of inputs. Note that in comparison to relatively primitive logic gates, neurons are tremendously complex and versatile logic devices, as they can integrate thousands of inputs from their dendrites, and process both temporally and spatially.
Logical Functions: Two Excitatory Inputs
Two strong excitatory inputs: logical OR. If either input is active, the neuron will be stimulated enough to fire.
Two weak excitatory inputs: logical AND. Both inputs must be active for the neuron to be stimulated enough to fire.
One weak, one strong excitatory input: logical IF. The strong input must be active for the weak input to exercise its effect. The activity of the neuron depends on the activity of the weak signal, but only if the strong signal is active.
Logical Functions: Mixed Inputs
Stronger inhibitory input, weaker excitatory input: logical IF-NOT. The inhibitory input will overwhelm the excitatory input if it is active. The neuronal activity depends on the weak excitatory input, but only if the strong inhibitory input is inactive.
Weaker inhibitory input, stronger excitatory input: logical MINUS. If the inhibitory input is sufficiently strong, it will cancel enough of the excitatory input to keep the neuron from firing.
Neurons are capable of expressing a tremendous array of logical functionality based on basic variations in strength and magnitude of synaptic inputs. However, unlike the traditional logic gates found in computer hardware, neurons are adaptable. The variables that control the functionality of neurons change based on internal and external environmental factors.
The cellular adaptation ultimately responsible for learning in the cerebral cortical is temporal memorization, across different time scales varying from milliseconds to years. Neurons are able to "memorize" information, preserving it over time through a variety of methods. This process is primarily electrochemical for short-term memory and structural for long-term and permanent memory.
Time Scale of Memory Operations
Electrical (1ms-100ms): ion flows due to transmission and basic neural information processing
Chemical (second-minute): changes in chemical balances and secondary messengers that affect receptions and ion channels in the cell membrane
Molecular (hour-day): molecular synthesis and gene expression leads to long-term modifications
Structural (day-year): structural changes to the cell itself, thus altering information processing, and to the membrane extensions (synapses and dendrites, connecting to other neurons and the outside)
Cellular memory includes any information that lingers following the immediate effect of an input-processing-output sequence. Besides providing the basis for long-term cortical information storage, cellular memory also allows neurons to transform temporal information into intensities. For instance, neurons can detect vibrations in the inner ear when oscillations occur at a certain frequency. This allows the brain to relate spatial and temporal information, which is critical for sequencing motor movements.