Connectionists: Brain calculus and brain logic: a firewall for sensory integration?

[Chi-Sang Poon]

Nonassociative learning as gated neural integrator and differentiator in
stimulus-response pathways

http://www.behavioralandbrainfunctions.com/content/pdf/1744-9081-2-29.pdf

 

Synopsis

What could drug addiction, the phantom pain experienced by amputees, and a
life-threatening respiratory condition called apneustic breathing have in
common? A new theory published in the online open-access journal Behavioral
and Brain Functions suggests that they all may be signs of brain calculus
and brain logic computations gone haywire. So says the paper’s lead author
Dr. Chi-Sang Poon, a scientist at the Harvard University-Massachusetts
Institute of Technology Division of Health Sciences and Technology. 

According to Poon, our brain is constantly bombarded with vast amounts of
sensory information that must be continuously sorted into actionable and
nonactionable items in order to prioritize. Such a complex mental task
involves sophisticated mathematical calculations like integral-differential
calculus and Boolean logic operations, which are basic to any
decision-making process.  But unlike the number crunching on digital
computers, the new theory proposes that our brain may be doing the math
automatically by using built-in neural circuitries capable of learning on
the spot. 

Such behavioral learning has long been thought to be a “dual process,” as
exemplified by the everyday experience of habituation to prolonged exposure
to fragrance and sensitization to recurrent shock and pain. Dr. Eric
Kandel’s pioneering work at Columbia University in the 1970s on neural
circuitries for habituation and sensitization in the sea slug Aplysia
resulted in a Nobel Prize in Physiology or Medicine in 2000 – but is such
quotidian behavioral learning really important to one’s well-being?

Then earlier in 2000, Poon’s research group discovered that the mammalian
brain displayed yet another mode of behavioral learning that had confounded
previous studies. They called this new behavior “desensitization,” in
contrast to sensitization and habituation. Their research demonstrated that
desensitization and habituation had similar “differentiator” effects on the
stimulus-response relationship, much like a “high-pass filter” in an audio
system. However, habituation was found to be turned on or off by the
stimulus itself, much like a Boolean toggle switch. Similarly, the effects
of sensitization were shown to be analogous to those of an “integrator” or
“low-pass filter,” with or without the Boolean on-off switching. 

These pivotal discoveries provided the pieces to the puzzle that inspired
Poon’s current theory in which habituation, sensitization, and
desensitization are the basic machinery for online calculus and logic
computations in the brain. The theory seems to bring these different
concepts together. In effect, behavioral learning is a form of brain
intelligence whereby integral-differential calculus and on-off Boolean
logics are used to filter incoming sensory signals in order to determine
continuously what needs attention and what doesn’t – which is our brain’s
way of telling “what’s hot and what’s not”. This so-called “sensory
firewall” allows the brain to relax and to economize its activities until
warning bells ring. It can also provide a fail-safe compensation when
sensory cues are distorted. A mistuning of the habituation or sensitization
components in the firewall could leave an individual either numbly
insensitive, as in “hearing without listening,” or excessively sensory
defensive, as in hysteria. Alternately, a breakdown of the desensitization
component could produce a sensory delusion. 

The effects of desensitization were discovered by Poon’s team while studying
the classic Hering-Breuer respiratory reflex. Here, the inspiratory drive is
slaked once the vagus nerves sense the lungs are inflated. (Try it yourself
by taking a deep breath and holding it. Momentarily, you will feel like you
want to exhale instead of inhale). This simple reflex triggers
inspiratory-to-expiratory phase switching, which is essential for
maintaining a cyclical respiratory rhythm. Poon’s group discovered that, in
animals whose vagus nerves are severed, a specific brainstem region in the
pons that is normally desensitized would steer the respiratory rhythm in
place of the vagus nerves. The pons seemed to act as a “phantom” or
surrogate for the vagus nerves – much like the phantom pain sensation
experienced by amputees. However, in this case, the compensatory action
provided an important respiratory fail-safe mechanism crucial for survival.
Indeed, classical experiments have shown that when both the vagus nerves and
the pons malfunction, an animal goes into an inhalation-only mode,
desperately trying to distend its lungs. This results in a life-threatening
neurological state called apneustic breathing. Poon and colleagues believe
this inspiratory-craving state is functionally similar to obsessive or
addictive behaviors, which may result when craving-inhibiting pathways in
reward-related brain regions are desensitized. If so, response
desensitization could be a new pattern for brain intelligence, and any
resulting errors in individual sensory systems may produce abnormalities
ranging from phantom pain to addictive behavior. 

In recent years, neuroscientists have been increasingly intrigued by the
idea that the human mind might be connected with the body’s environment
through the construction of certain internal models. This was hinted by the
seventeenth-century French philosopher and mathematician René Descartes. If
Poon and colleagues are correct, the sensory firewall mediated by
nonassociative learning may be the gatekeeper of the internal models that
govern sensory integration in the brain.