Reward prediction error and fMRI: do they go together?

New paper: Bissonette GB, Gentry RN, Padmala S, Pessoa L, Roesch MR. Impact of appetitive and aversive outcomes on brain responses: linking the animal and human literatures. Front Syst Neurosci. 2014 Mar 4;8:24. eCollection 2014.

With Matt Roesch and folks in his lab, we have a new paper reviewing the animal and animal literatures on appetitive and aversive processing. But what I want to discuss here is one issue that became apparent when we started collecting some pilot fMRI data. Whereas work in the rat has shown prediction errors in a few places in the brain, some fMRI work has shown quite widespread signals. This might be true but one possible confound that we find in many papers in the literature is that regions that simply respond to the US stimulus could be incorrectly described as showing prediction errors because of the way the fMRI analysis is done. This is explained in Fig. 6 of our paper and in the figure below (thanks to Brenton McMenamin).

The problem is that when a regressor modeling the US (say, reward) is also introduced in the model in addition to the prediction error regressor, it can absorb variance in a way that what is left is actually how a prediction error looks like. In that way, the region will appear to show a prediction error simply because it responds to reward itself.

I have no idea how prevalent this problem is, and it is at times even unclear how the data were modeled. (In fact, as an aside, it is at times very surprising how people don’t explain what they do in their papers.) Some people have talked about this problem but again I’m not sure how much people are taking this into account.

fig6_v2 (1)

Understanding brain networks and brain organization: new paper

New paper to appear in Physics of Life that will come with commentary (including by Dani Bassett, Barry Horwitz, Claus Hilgetag, Vince Calhoun, Michael Anderson, Evan Thompson, Franco Cauda, among others).

A lot has been written about brain networks, especially say after 2005. In Chapter 8 of my book The Cognitive Emotional Brain I wrote about this from the perspective of understanding structure-function mappings (what do regions do? what do networks do?). In a paper in press in Physics of Life, I update some of my evolving thoughts on this question. Some of the newer points are:

  • Is brain architecture really small world?  Cortical connectivity seems too dense.  But an important ingredient of small-world organization — the existence of non-local connections, especially long-range ones – is clearly present. Although they appear to be relatively weak, long-range connections play a major role in the cortical network.
  • The mapping from network (as a set of regions) to function is not one-to-one. For instance: Menon, Uddin, and colleagues suggest that a salience network involving the anterior insula and the anterior cingulate cortex “mediates attention to the external and internal worlds”. They note, however, that “to determine whether this network indeed specifically performs this function will require testing and validation of a sequence of putative network mechanisms…” I argue that a network’s operation will depend on several more global variables, namely an extended context that includes the state of several “neurotransmitter systems”, arousal, slow wave potentials, etc. In other words, a network that is solely defined as a “collection of regions” is insufficient to eliminate the one-to-many problem observed with brain regions (such as the amygdala being involved in several functions).
  • Cortical myopia (echoing points by Parvizi, 2009). Large-scale analyses and descriptions of brain architecture suggest principles of organization that become apparent when information is combined across many individual studies. Unfortunately, most of these “meta” studies are cortico-centric – they pay little or no attention to subcortical connectivity. This paints a rather skewed view of brain organization. For example: if one considers “signal communication” as proposed by Sherman (see figure), cortical communication might go via thalamus (including pulvinar), flipping the traditional view.

    Scheme by Sherman SM. The thalamus is more than just a relay. Current Opinions in Neurobiology. 2007;17:417-22.

    Scheme by Sherman SM. The thalamus is more than just a relay. Current Opinions in Neurobiology. 2007;17:417-22.

  • Evolution. Related to the previous point, I suggest that to understand the contributions of subcortical connectivity, we need to consider the evolution of the brain. For example: a cortico-centric framework is one in which the “newer” cortex controls subcortical regions, which are typically assumed to be relatively unchanged throughout evolution. Instead, I suggest that cortex and subcortex change in a coordinated fashion.
  • The importance of weak connections. I critique a central component of the “standard” network view, which goes something like this: “network states depend on strong structural connections; conversely, weak connections have a relatively minor impact on brain states.” My contention is that weak connections are much more important.


Reference: Parvizi J. Corticocentric myopia: old bias in new cognitive sciences. Trends in Cognitive Sciences. 2009;13:354-9.