Aversive learning enhances perceptual and cortical discrimination

Li, W, Howard, J.D., Parrish, T.B., & Gottfried, J.A. (March 28, 2008). Aversive learning enhances perceptual and cortical discrimination of indiscriminable odor cues. Science, Vol. 319, 1842-1844.

With this study, the authors explored the impact of aversive conditioning on olfactory discrimination. While most conditioning studies examine the acquisition of new behavioral responses (CR) to formerly benign stimuli presentations, this examined how associative learning can actually alter the perceptual processing of the conditioned stimulus (CS) itself. Following a conditioning regimen, behavioral accuracy for distinguishing by smell between a previously indistinguishable pair of molecules (CS+) rose by more than a factor of 2, exceeding both chance and preconditioning performance. Interestingly, following conditioning, no improvement in distinguishing between the unconditioned control pair (CS-) was witnessed, indicating that these effects are specific to the CS+. After conditioning, reorganization of neural coding was also observed in the posterior piriform cortex, where neural representations of odor identity are maintained. This may shed new light on anxiety disorders which are characterized by exaggerated sensory sensitivity and hypervigilance, potentially self-reinforcing patterns.

Human amygdala activation during conditioned fear acquisition and extinction

LaBar, K.S., Gatenby, J.C., Gore, J.C., LeDoux, J.E., & Phelps, E.A. (May 1998). Human amygdala activation during conditioned fear acquisition and extinction: a mixed-trial fMRI study. Neuron, Vol. 20, 937-945.

The amygdala is believed to be a key component in a network mediating survival functions by coordinating behavioral plans of action based on the integration of exteroceptive and interoceptive information. The amygdala, in particular, has been thought to be the structure responsible for detecting and reacting to potentially threatening environmental stimuli through classical conditioning learning.

However, attempts to investigate amygdala function in humans has produced inconsistent results, with failures to notice increased amygdala blood flow in PET being most surprising. Difficulties may stem from the small size and troublesome subcortical placement of the structure, and that amygdalar responses are relatively transient to discrete cues, have low spontaneous neuronal firing rates, and exhibit marked habituation (gradual signal intensity reduction). This study attempted to overcome these difficulties by using a mixed-trial fMRI design. This time, results successfully showed amygdala/periamygdaloid cortex involvement during both conditioned fear acquisition and extinction, biased towards the right hemisphere in both cases.

Previous lesion studies have shown the integrity of the amygdala is required for expression of learned conditioned fear associations. However, the temporal pattern of amygdala activity (greatest during early acquisition and early extinction, and degrading over time) suggests that this activity may only partially underlie expression. The paper offers a hypothesis that the observed activation may be related to encoding the emotional meaning of the conditioned stimulus. This is consistent with with the amygdala activity witnessed during the initial stages of learning (when stimulus is novel) and during early extinction (when the meaning of the stimulus has changed).

Brain systems mediating aversive conditioning

Buchel, C., Morris, J., Dolan, R.J., & Friston, K.J. (May 1998). Brain systems mediating aversive conditioning: an event-related fMRI study. Neuron, Vol. 20, 947-957.

Classical conditioning refers to a type of associative learning whereby a previously neutral stimulus (CS) comes to elicit a behavioral response by being paired with an aversive unconditioned stimulus (US). This study implemented a human classical conditioning paradigm in which images of faces (CS) were paired with an aversive tone (US). To assess which areas of the brain were related to conditioning, event-related fMRI responses were compared between presentation of conditioned stimuli (CS+) and unconditioned stimuli (CS-) after skin conductance indicated the conditioning regimen was completed successfully. To be more accurate, CS- was compared to the occasional trials of CS+ which were not followed by a tone.

Unequivocal differential responses were found in two cortical areas: the anterior cingulate cortex (ACC) and the anterior insula. These structures receive input from the amygdala, which also shows interesting activation patterns during this paradigm. The lateral amygdala in particular shows time-dependent neural responses, with higher than baseline responses at first but habituating over time. A possible explanation is a negative feedback loop, an analgesia kicked off by the amygdala and mediated by endogenous opioids which leads to reduced conditioning over time. Finally, differential activation was also witnessed in the red nucleus together with premotor structures, characteristic of response expression.

Happy 50th post!!! And happy bissextile day!!!

dlPFC promotes LTM formation via WM organization

Blumenfeld, R.S. & Ranganath, C. (January 18, 2006). Dorsolateral prefrontal cortex promotes long-term memory formation through its role in working memory organization. The Journal of Neuroscience, 26, 3, 916-925.

This paper speculates that the different regions of the prefrontal cortex contribute differentially to working memory (WM). Specifically, they argue that the dorsolateral prefrontal cortex (dlPFC) is recruited during tasks requiring organization of items active in WM, while the ventrolateral prefrontal areas (vlPFC) are involved in WM maintenance, or simply holding items in short-term memory.

The study begins behaviorally by showing that overall memory is increased during tasks which require organization by re-ordering (in comparison to tasks which are merely rehearsals). Single items encoded on re-ordering trials were significantly more likely to be judged as "remembered" than from rehearsal trials. And organization of items during re-ordering trials resulted in much higher recollection together, suggesting the strengthening of associative links between items in these organization trials. Then the paper shifts to the fMRI data to attempt to implicate the dlPFC specifically. Their evidence: the dlPFC showed increased activation during the delay (encoding) period of re-ordering trials relative to rehearse trials.

However, one very possible alternative explanation is the effect of nonspecific factors correlated with task difficulty. Reorder and rehearse trials differ in a number of ways, difficulty being the most significant. Perhaps the increased attention required in the re-ordering trials led to changes in behavioral results and a general increase in activation in the dlPFC (and elsewhere). This raises concerns that the activity in the dlPFC could be attributed to processes other than organization.

Changes in Hippocampi of Taxi Drivers

Maguire et al. (April 11, 2000). Navigation-related structural change in the hippocampi of taxi drivers. PNAS, Vol. 97, No. 8, 4398-4403.

This correlational study showed taxi drivers had a significantly greater volume in the posterior hippocampus, whereas control subjects showed greater volume in the anterior hippocampus. Authors interpret the results as evidence for the relative redistribution of gray matter in the hippocampus in response to the occupational need to store increasingly detailed spatial representations. The amount of time spent as a taxi driver was found to be correlated with the amount of volume in the right posterior hippocampus, suggesting that while mental spatial maps are likely to be stored in the posterior hippocampus and necessitate progressive structural changes over time, the left hippocampus may participate in spatial navigation and memory differently from the right.

Differential processing of objects under various viewing conditions in the human LOC

Grill-Spector, K. et al. (September 1999). Differential processing of objects under various viewing conditions in the human lateral occipital complex. Neuron, Vol. 24, 187-203.

This study used fMRIa techniques to investigate the brain's object-selective regions (namely the lateral occipital complex). fMRIa assumes that a group of neurons will respond to repeated presentations of a stimulus with attenuated responses. This signal reduction is presumably the result of neural fatigue from repeated exposure. Thus, "adaptation" data can be used to identify which types of stimulus are effectively treated identically by a certain region of the brain. This is especially interesting in structures further up the processing hierarchy where lower-level transformations have likely given way to more abstract, general representations of the stimuli. Researchers can then explore what properties of objects are preserved and which are transformed to a canonical representation by the time signals converge on a specific region.

First, the study sought to understand how long adaptation effects last. Time durations as long as 8 sec between matched stimuli still elicited amplitude reductions, establishing that adaptation has a fairly long-lasting effect.

Secondly, the study set out to examine which object properties were invariant within the LOC. The results indicate that the LOC is less sensitive to changes in size and position, compared to changes induced by illumination and viewpoint (rotation). In other words, it seems the LOC receives visual input which has been normalized for size and position.

fMRIa

Krekelberg, B, Boynton, G.M., & van Wezel, R.J.A. (2006). Adaptation: from single cells to BOLD signals. Trends in Neuroscience.

Functional magnetic resonance imaging adaptation (fMRIa) is an increasingly popular method which takes advantage of the brain changes which occur in response to long exposure to some evocative stimulus. If Stimulus 1 (S1) excites a certain neuronal population, repeated exposure to S1 will result in subsequently attenuated responses. This may be due to neural fatigue (i.e. the more a neuron fires, the more its subsequent responses will be reduced) or may be due to coupled hemodynamic processes. However, when S1 is followed by a unique stimulus, S2, the response amplitudes should not be attenuated as a fresh sub-population of neurons is excited. Using this technique can allow researchers to determine if the same or unique neuronal groups are involved in processing two stimuli. This paper goes on to describe the utility of the technique in examination of the visual system, particularly orientation, motion, and face detection. It also stresses the importance of adaptation timescale in experimental design.

Retinotopy and Functional Subdivision of Human Areas MT and MST

Huk, A.C. et al. (August 15, 2002). Retinotopy and Functional Subdivision of Human Areas MT and MST. The Journal of Neuroscience, Volume 22, Number 16, 7195-7205.

Much research in neuroscience has begun with animal studies (and invasive techniques) and only later been continued with human subjects (with predominantly non-invasive methodologies). This paper discusses the attempts to reconcile the areas of visual cortex responsible for detecting visual motion in the macaque and the human. In particular, two sub-divisions of the dorsal superior temporal sulcus (STS) are well-studied in the macaque: the middle temporal (MT) and the medial superior temporal (MST) visual areas. The MT is characterized by a distinguishable retinotopic map and a coarse-grained small receptive field, whereas the MST is just the opposite. This experiment was designed to recognize areas displaying these characteristics in the human MT+ or V5 area, thought to be homologous to the macaque STS, using fMRI. Subregions of the human MT+ were tentatively identified, bridging the gap between animal and human research.

Prediction of immediate and future rewards differentially recruits cortico-basal ganglia loops

Tanaka, S.C. et al. (August 2004). Prediction of immediate and future rewards differentially recruits cortico-basal ganglia loops. Nature Neuroscience, Volume 7, Number 8, 887-893.

Curiously, lesions in the nucleus accumbens in rats result in a tendency to choose small immediate rewards over larger future rewards. And low activity in the central serotonergic system is also associated with impulsive behavior in humans. This article argues that the lateral orbit-frontal cortex (OFC) takes on the role of predicting immediate rewards, while the dorsolateral prefrontal cortex (dlPFC), dorsal pre-motor cortex (dMC), and inferior parietal cortex (IPC) are involved in prediction of future outcomes. Therefore, different sub-loops of the cortico-basal ganglia network are specialized for reward prediction at different time scales. The last piece of the puzzle, the dorsal raphe nucleus, is thought to use serotonin influence to control the effective time scale of reward prediction, allowing flexible selection of a relevant time-scale appropriate for the task at the time of decision-making.

Using human brain lesions to infer function

Rorden, C. & Karnath, H. (October 2004). Using human brain lesions to infer function: a relic from a past era in the fMRI age? Nature Reviews, Vol. 5, 813-819.

Patients with brain lesions have historically provided neuroscience with momentous insights into brain function. However, with the advent of non-invasive in vivo imaging techniques such as CT, MRI, fMRI, DTI, and the like, researchers are questioning the role of lesion methods going forward. This article argues that lesions studies will continue to fill a unique niche in the future, especially in combination with new imaging protocols: "The power of cognitive neuroscience comes from using convergent tools to investigate the same theoretical question."

Transcranial magnetic stimulation and cognitive neuroscience

Walsh, V. & Cowey, A. (October 2000). Transcranial magnetic stimulation and cognitive neuroscience, Nature Reviews, Volume 1, 73-79.

Transcranial magnetic stimulation (TMS) is an investigative tool used in neuroscience to transiently interfere with brain functions, temporarily interrupting normal brain activity in a restricted region of the brain. This non-surgical technique introduces random activity as disorder into the information processing system, disrupting cognitive task performance. When coupled with a neuroimaging technique such as fMRI or EEG, TMS has proven very useful in experimentation to elucidate brain function.

Throughout the history of neuropsychology, patients with brain injuries ranging from mild to major have provided critical insights into the way the brain operates. As a mechanism of interference, TMS is unique in that it can be used to create 'virtual lesions', short-lived and reversible. Real lesioned brains will have undergone months if not years of neural reorganization following an accident to compensate for the deficit. Since TMS interference is typically not prolonged, no such opportunities exist for long-term reorganization processes to kick in. While this may seem like a poor approximation of classical lesioning, researchers believe that studying these types of temporary lesions may in fact be more useful to scientific examination without the introduction of the brain's compensatory coping strategies.

In addition to pure interference studies, the effects of TMS at the primary site of application have been shown to correspond very well with activation produced by self-induced behavior. For example, in one study activation was presented to the motor cortex above the motor threshold, creating behavior in the arm. When subjects were asked to reproduce the same arm movement voluntarily, there was great similarity between the two compared brain activations. However, one restriction of non-surgical TMS is that stimulation is limited to superficial cortical regions.

What does fMRI tell us about Neuronal Activity?

Heeger, D.J. & Ress, D. (February 2002). What does fMRI tell us about Neuronal Activity?, Nature Reviews, Volume 3, 142-151.

The linear transform model refers to a fundamental assumption guiding the analysis of fMRI studies, namely that fMRI signal is approximately proportional to a measure of local neural activity. Clearly, the ultimate success of this research depends on a clear relationship between the fMRI signal and the underlying neuronal activity. However, some scientists believe reasons exist to be skeptical of this pervasive assumption. Heeger and Ress provide a review of arguments on both sides, concluding that this model is a reasonable and useful approximation, but only for some recording sites, in some brain areas, and using certain experimental protocols.

This is Your Brain on Politics

Iacoboni, Freedman, & Kaplan. (November 11, 2007). This Is Your Brain on Politics, The New York Times.

Easily one of the most bogus articles ever published on topics related to neuroscience. In short, the article claimed that it is possible to read the minds of people by examining their brain activity a la fMRI and glean insight into their cognitive reactions to American presidential candidates. The most fallacious reasoning employed by the authors was the assumption that activity in certain brain regions can have a 1:1 mapping to mental states such as anxiety, disgust, equivocation, empathy, and interest, e.g. the amygdala was active so they must be anxious with regard to this candidate. Additionally, the article was not peer reviewed, did not provide any scientific detail which could be used to evaluate the conclusions drawn, and only "studied" 20 subjects.

Luckily, there were some expedient retorts by subject matter experts, but probably not before millions of readers were led to believe that this is acceptable science and produced reasonable logical conclusions. Letters to the Editor, Nature

Selective Effects of Ritalin in ADHD

Vaidya, C.J., Austin, G., Kirkorian, G., Ridlehuber, H.W., Desmond, J.E., Glover, G.H., & Gabrieli, J.D.E. (November 1998). Selective effects of methylphenidate in attention deficit hyperactivity disorder: A functional magnetic resonance study, Neurobiology, 95, 14494-14499.

ADHD is the most common developmental disorder of childhood and has been associated with such adverse life outcomes as lower educational and occupational achievement, as well as increased risk for various disorders in adulthood. Evidence suggests that ADHD is characterized by dysfunction in transmission of dopamine to the frontal lobes and striatal (basal ganglia) structures of the brain since ADHD symptoms typically respond favorably to stimulant medications (e.g. methylphenidate) that release and inhibit reuptake of dopamine in these regions. However, prior to this study, there was no direct evidence to indicate differences in dopaminergic modulation between ADHD and normal children. Therefore, the purpose of this study was to see: (a) how performance in response inhibition tasks differs between ADHD children and normal children, (b) how administration of methylphenidate improved performance in response inhibition tasks in ADHD children in comparison with normal children, (c) how baseline frontal-striatal function differ in ADHD and control children, and (d) how methylphenidate modulates frontal-striatal function differently in ADHD and control children. Since inhibition of motor responses is known to depend on the integrity of both frontal and striatal structures, functional magnetic resonance imaging (fMRI) was used to image the frontal lobes, as well as the head of the caudate nucleus and the anterior portion of the putamen during response inhibition tasks to assess the four aforementioned points.

Each subject in the study was presented with various computer-generated stimuli and instructed to respond with a button-press on a hand-held joystick. In one block, they were told to respond to all stimuli presented. In another block, they were told to respond to all stimuli except for one (inhibition). Their behavioral performance was measured by the percentage of errors committed during the task. At the same time, activation of certain brain regions was gathered during the trials via fMRI. With the different independent variables and levels, the overall design looks something like the following table for the "stimulus-controlled" task (for brevity, I will neglect to mention the "response-controlled" task and results):



The results can be condensed as follows: (a) The ADHD group made more errors than the control group on the response inhibition task. (b) Both the ADHD and the control groups showed significant improvements on the task with the administration of MPH. (c) Baseline striatal activation was shown to be lower in ADHD subjects than in the control group. (d) Administration of MPH increased striatal activation in ADHD subjects but decreased striatal activation in control subjects.

The results support the hypothesis that there are indeed differences between ADHD children and normal children, both in performance on response inhibition tasks and in their striatal activation in the absence of stimulant medication. And further, when administered stimulant medication, ADHD children showed different functional reactions to the medication in the striatum than their control group counterparts. These are important findings, especially given the neural specificity of the results.