NeuroImage 62 (2012) 1216–1220

Contents lists available at SciVerse ScienceDirect

NeuroImage

j ourna l homepage: www.e lsev ie r.com/ locate /yn img

Review

The future of fMRI in cognitive neuroscience

Russell A. PoldrackDepartments of Psychology and Neurobiology and Imaging Research Center, University of Texas, Austin, TX 78759, USA

E-mail address: poldrack@mail.utexas.edu.

1053-8119/$ – see front matter © 2011 Elsevier Inc. Aldoi:10.1016/j.neuroimage.2011.08.007

a b s t r a c t

a r t i c l e i n f o

Article history:Accepted 5 August 2011Available online 11 August 2011

Keywords:fMRICognitive neuroscienceMeta-analysisOntologyConnectivity

Over the last 20 years, fMRI has revolutionized cognitive neuroscience. Here I outline a vision for what thenext 20 years of fMRI in cognitive neuroscience might look like. Some developments that I hope for includeincreased methodological rigor, an increasing focus on connectivity and pattern analysis as opposed to“blobology”, a greater focus on selective inference powered by open databases, and increased use ofontologies and computational models to describe underlying processes.

l rights reserved.

© 2011 Elsevier Inc. All rights reserved.

Contents

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1216Methodological rigor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1216From blobology to pattern analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1217Selective inference . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1218Open science and data aggregation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1218Describing mental processes: ontologies and computational models . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1219Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1219Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1219References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1219

Introduction

When I began graduate school for cognitive psychology in 1989, Iwas unaware of the ongoing race to develop functional magneticresonance imaging (fMRI). It may come as no surprise to those whoknow my writings that when I first encountered imaging studies as agraduate student, I was highly skeptical that it could inform us aboutmental function. I was, in this sense, in good company; in those days, itwas not uncommon to hear conversations between cognitivepsychologists likening fMRI to trying to understand how an engineworks by measuring the temperature of the exhaust manifold. When Imoved to Stanford in 1995 to start my postdoc, I didn't actually go thereto do fMRI, but circumstances led me to get involved in the imagingwork that was ongoing in John Gabrieli's lab, and my career as aneuroimager was born.

Mythoughtsbelowabout the futureof fMRI in cognitiveneurosciencewould be better characterized as hopes rather than predictions. Despitewhat I see as serious fundamental problems in how fMRI has been andcontinues to be used in cognitive neuroscience, I think that the last fewyears havewitnessed a number of encouraging newdevelopments, and Iremain very hopeful that fMRI will continue to provide us with usefulinsights into the relation between mind and brain. In the foregoing, Ioutline what I see as the the most promising new directions for fMRI incognitive neuroscience, with an obvious bias towards the some ofdirections that my own research is currently taking.

Methodological rigor

Foremost, I hope that in the next 20 years the field of cognitiveneuroscience will increase the rigor with which it applies neuroim-aging methods. The recent debates about circularity and “voodoocorrelations” (Kriegeskorte et al., in press; Vul et al., 2009) havehighlighted the need for increased care regarding analytic methods.Consideration of similar debates in genetics and clinical trials led

1217R.A. Poldrack / NeuroImage 62 (2012) 1216–1220

(Ioannidis, 2005) to outline a number of factors thatmay contribute toincreased levels of spurious results in any scientific field, and thedegree to which many of these apply to fMRI research is rathersobering:

• small sample sizes• small effect sizes• large number of tested effects• flexibilty in designs, definitions, outcomes, and analysis methods• being a “hot” scientific field

Some simple methodological improvements could make a bigdifference. First, the field needs to agree that inference based onuncorrected statistical results is not acceptable (cf. Bennett et al., 2009).Many researchers have digested this important fact, but it is stillcommon to see results presented at thresholds such as uncorrectedpb .005. Because such uncorrected thresholds do not adapt to the data(e.g., the number of voxels tests or their spatial smoothness), they arecertain to be invalid in almost every situation (potentially being eitheroverly liberal or overly conservative). As an example, I took the fMRIdata from Tom et al. (2007), and created a random “individualdifference” variable. Thus, there should be no correlations observedother than Type I errors. However, thresholding at uncorrected pb .001andaminimumcluster size of 25 voxels (a commonheuristic threshold)showed a significant region near the amygdala; Fig. 1 shows this regionalongwith a plot of the “beautiful” (but artifactual) correlation betweenactivation and the random behavioral variable. This activation was notpresent when using a corrected statistic. A similar point was made in amore humorous way by Bennett et al. (2010), who scanned a deadsalmonbeingpresentedwith a social cognition task and foundactivationwhen using an uncorrected threshold. There are now a number of well-established methods for multiple comparisons correction (Poldrack etal., 2011), such that there is absolutely no excuse to present results atuncorrected thresholds. The most common reason for failing to userigorous corrections formultiple tests is that with smaller samples thesemethods are highly conservative, and thus result in a high rate of falsenegatives. This is certainly a problem, but I don't think that the answer isto present uncorrected results; rather, the answer is to ensure that one'ssample is large enough to provide sufficient statistical power to find theeffects of interest.

Second, I have become increasingly concerned about the use of“small volume corrections” to address the multiple testing problem.The use of a priori masks to constrain statistical testing is perfectly

rand

om c

ovar

iate

2.5

2

1.5

1

0.5

0

-0.5

-1

-1.5

-2

Fig. 1. An example of artifactual activation that survives a heuristic statistical correction. T25 voxel extent threshold) for the correlation between activation and a randomly generatedthe map and the randomly generated variable, showing that completely spurious data cancircular data analysis.

legitimate, but one often gets the feeling that the masks used for smallvolume correction were chosen after seeing the initial results(perhaps after a whole-brain corrected analysis was not significant).In such a case, any inferences based on these corrections are circular andthe statistics are useless. Researchers who plan to use small volumecorrections in their analysis should formulate a specific analysis planprior to any analyses, and only use small volume corrections that wereexplicitly planned a priori. This sounds like a remedial lesson in basicstatistics, but unfortunately it seems to be regularly forgotten byresearchers in the field.

Third, the field needs to move toward the use of more robustmethods for statistical inference (e.g., Huber, 2004). In particular,analyses of correlations between activation and behavior acrosssubjects are highly susceptible to the influence of outlier subjects,especially with small sample sizes. Robust statistical methods canensure that the results are not overly influenced by these outliers,either by reducing the effect of outlier datapoints (e.g., robustregression using iteratively reweighted least squares) or by separatelymodeling data points that fall too far outside of the rest of the sample(e.g., mixture modeling). Robust tools for fMRI group analysis areincreasingly available, both as part of standard software packages(such as the “outlier detection” technique implemented in FSL:Woolrich, 2008) and as add-on toolboxes (Wager et al., 2005). Giventhe frequency with which outliers are observed in group fMRI data,these methods should become standard in the field. However, it's alsoimportant to remember that they are not a panacea, and that itremains important to apply sufficient quality control to statisticalresults, in order to understand the degree to which one's resultsreflect generalizeable patterns versus statistical figments.

From blobology to pattern analysis

The foregoing methodological issues have taken on particularimportance due to the strong focus of the field on localization offunction; the goal of finding blobs in a specific region can driveresearchers into analytic gymnastics in order to find a significant blobto report. However, for the last few years the most interesting andnovel research has focused on understanding patterns of activationrather than localized blobs. The appreciation of patterns is happeningat multiple scales. At the systems (whole-brain) scale, themodeling ofconnectivity and its relation to behavior continue to grow. Oneimpediment to the analysis of connectivity with fMRI is the lack of a

-150 -100 -50 0

fmri signal50 100 150 200

r=0.781

he left panel shows the thresholded activation map (pb .001, voxelwise uncorrected,variable. The right plot shows the correlation between activation in the peak voxel fromproduce seemingly compelling results when heuristic corrections are combined with

1218 R.A. Poldrack / NeuroImage 62 (2012) 1216–1220

standardized set of approaches for connectivity modeling; with somany different approaches being used across the field, it can bedifficult to understand how the results relate to one another. Therecent work by Smith et al. (2011) examining the relative power ofdifferent modeling approaches using simulated data is a good starttoward better understanding the relative strengths andweaknesses ofthese different methods, and I envision that these kinds ofmethodological “shoot-outs” will eventually help determine whichmethods are most optimal for which questions. I think that the jury isstill out on how well fMRI can ever characterize neuronal connectiv-ity; as we outlined in Ramsey et al. (2010), there are a number offundamental challenges in using fMRI to characterize causal in-teractions between brain regions. Further work combining neuro-physiology and fMRI (e.g., David et al., 2008) will be essential to fullyunderstand the potential and limitations of fMRI for modeling of theunderlying neuronal connectivity.

At a more local level, the use of analyses that focus on patterns ofactivation over a region rather thanmean activation in that region hasbecome very popular in the last few years (e.g., Kriegeskorte et al.,2006, 2008; Norman et al., 2006), and I expect that this will continue.It is now clear that information can be coded in fMRI data eitherthrough local changes in mean activation or through changes in localpatterns of activity, and that analysis of only the mean activation(which has been the standard over the last 20 years) may bemissing asignificant part of the picture. For example, we recently comparedresults from closely-matched univariate and multivariate (pattern-information) analyses for a decision-making task, and found manyregions in the cortex where the multivariate test was significantlymore sensitive to task-relevant information than the univariateanalysis, with a smaller set of regions showing the opposite pattern(Jimura and Poldrack, in press). I think that one approach that hasparticularly strong appeal is representational similarity analysis(Kriegeskorte et al., 2008), in which the patterns of activity in fMRIdata are related to behaviorally-relevant variables such as stimulusfeatures or task manipulations. For example, we (Xue et al., 2010)recently used this approach to show that memory is better for itemswhose patterns of activity are more similar across study episodes,which has implications for cognitive theories of memory. I think thatthis approach has great potential to provide much stronger links fromneuroimaging data to cognitive theories (which often make detailedpredictions about patterns across stimuli) and computational models,which will help provide a stronger basis for relating brain activity andmental function.

Selective inference

Neuroimaging as it is used by most cognitive neuroscientists facesa fundamental problem, which is evident from even a cursory readingof the literature: Within different subfields in cognitive neuroscience,very different functions can be assigned to the same structure. Theanterior cingulate may be the poster child for this heterogeneity, withproposed functions as various as conflict monitoring, interoception,pain, autonomic regulation, effort, and consciousness. This becomesparticularly problematic when researchers are aware of only one ofthose literatures and attempt to make reverse inferences based onactivation in a region (e.g., “the anterior cingulate is active, thus thesubject must be experiencing conflict”) (Poldrack, 2006).

If the goal of cognitive neuroscience is to map function ontostructure, then this is a serious problem, and the standard approach toneuroimaging does not provide any way to solve it. How can weidentify what the basic function is that a region or network subserves?As I argued recently (Poldrack, 2010), I think that the answer is tomove toward methods that allow us to identify selective associationsbetween functions and structures. This requires that we ask notsimply whether we can find a region that is engaged by a particularmental process, but whether one can find a region that is engaged

selectively, such that activation of the region is actually predictive ofthe mental process. The tools from machine learning that haverecently been brought to bear on fMRI data (Norman et al., 2006;Pereira et al., 2009; Poldrack et al., 2009) can provide importantleverage on this issue, by providing a principled way to assess thepredictive power of activation and thus allowing a quantitativeassessment of the strength of any reverse inference (Poldrack, 2008).

In addition to these tools, however, we need data sets that arebroad enough to allow the assessment of any particular functionagainst many other functions in order to assess selectivity andunderstand the larger-scale structure of mind-brain relationships. Todate, the only databases large enough to support such meta-analysisare based on activation coordinates mined from papers, an approachpioneered by the BrainMap group (Laird et al., 2005). We (Yarkoni etal., in press) have recently developed a coordinate-based database(http://www.neurosynth.org) that allows mining of full text inrelation to brain activation. This allows the creation of maps reflectingthe power of “forward inference” (i.e., inferring the presence ofactivation in each voxel given the presence of a particular term in thepaper) and “reverse inference” (i.e., inferring the presence of a term inthe paper given activation in a voxel). The striking insight to comefrom analyses of this database (Yarkoni et al., in press) is that someregions (e.g., anterior cingulate) can show high degrees of activationin forward inference maps, yet be of almost no use for reverseinference due to their very high base rates of activation across studies(e.g., some voxels in the anterior cingulate are active inmore than 25%of papers). Tools like this, which combine text mining withneuroimaging data, have the potential to increase our understandingof selective associations between mental function and brain activity.

Open science and data aggregation

The need to integrate information across many different taskparadigms and psychological domains highlights the critical role ofeffective data sharing across research groups, as it would beexceedingly difficult for an individual research group to collect thelarge amounts of data needed for such analyses. While thecoordinate-based meta-analytic approaches mentioned above willbe useful for many questions, they will undoubtedly fall short insome cases, leading to a need for the analysis of raw fMRI data.When the fMRI Data Center (fMRIDC) was started in 2001 with thegoal of sharing raw fMRI data (Van Horn et al., 2001), there waswidespread skepticism and opposition regarding the utility of datasharing. Although there were certainly some missteps in theexecution of the fMRIDC project, this reaction largely reflected thefact that the project was ahead of its time, and there was a lack of asocial consensus in favor of data sharing. A number of developmentssuggest that this has changed, such that the field is now ready forlarge-scale data aggregation. First, there is an incipient “openscience” revolution in scientific publishing, examples of which canbe seen in the Frontiers and Public Library of Science (PLOS)journals. In parallel, there is growing interest across science in thevalue of open access to large datasets. The powerful insights thathave arisen from data sharing in other fields (such as genomics,where every genome sequenceis shared via GenBank) have providedadditional proof of concept for the power of large shared databases.Together, these developments suggest the possibility of large-scaleeffective data sharing in the future. Groups such as the ScienceCommons (http://sciencecommons.org/) are developing new modelsfor open access to publications, data, and scientific tools. Theseefforts have been heavily inspired by the open source softwaremovement, which has demonstrated the effectiveness of “crowd-sourcing” (Doan et al., 2011).

The recent development and publication of large-scale restingstate fMRI datasets by the 1000 Functional Connectomes Project (FCP)and International Neuroimaging Data Initiative (INDI) have

1219R.A. Poldrack / NeuroImage 62 (2012) 1216–1220

demonstrated the feasibility of sharing large fMRI datasets in a freeand open manner (Biswal et al., 2010). However, sharing of task fMRIdata remains much more challenging due to the need for additionalmetadata describing the task. We have recently begun to address thischallenge via the OpenfMRI Project (http://www.openfmri.org).Although currently very small in comparison to other databasessuch as FCP/INDI, this project has developed a framework for therepresentation of task-based fMRI metadata that should make thesharing of these data significantly easier. As tools for data sharing aredeveloped and integrated with tools for data analysis (i.e., a “submit”button within the software package that would automatically uploadthe data to a shared database), I hope that we will see that sharing offMRI data will become increasingly common. I hope that as suchdatabases grow, they will provide a unique resource for testing thepower of reverse inference across a much larger set of tasks.

The utility of such large shared datasets also depends on sufficientcomputing resources to process hundreds or thousands of subjects'worth of fMRI data. The development of high-performance computingclusters based on commodity hardware has begun to provide suchresources; for example, using the clusters at the Texas AdvancedComputing Center we regularly run jobs that utilize more than 4000processors, allowing us to complete analyses in minutes that wouldhave previously taken months to finish. As cognitive neuroscienceresearch becomes increasingly reliant on analysis of very largedatasets, the use of such high-performance computing systems willbecome critical, and effective use of these resources will requirecognitive neuroscience researchers to interact even more closely withcomputer scientists and informaticians in order to take the bestadvantage of them.

Describingmental processes: ontologies and computational models

When cognitive neuroscience researchers use fMRI to map mentalprocesses onto brain structure, the processes that are being mappedare generally described in an informal way in the text of thepublication. While this has long been viewed as adequate, it makesthe aggregation of hypothesis-relevant data very difficult. In otherareas of bioscience such as genomics, the use of ontologies (formaldescriptions of terms and their relationships) to annotate data has ledto incredible progress in mapping of biological functions onto geneticstructure (Bard and Rhee, 2004), and it is likely that ontologies willplay an equally important role for cognitive neuroscience (Bilder et al.,2009; Yarkoni et al., 2010). If we are to make good use of the data thatare shared across research groups, these data will need to be deeplyannotated using an ontology of mental function, in order to then askwhich proposed associations between mental processes and brainsystems are supported across multiple task domains. Development ofsuch ontologies is very challenging because psychologists do notagree about much of the underlying mental architecture; however,capturing both the agreements and disagreements is fundamental ifwe wish to ultimately use brain imaging to understand theorganization of the mind. The Cognitive Atlas project (http://www.cognitiveatlas.org) (Poldrack et al., submitted for publication) hasbegun to develop such an ontology, which will serve as a basis forannotation in the OpenfMRI.org data sharing project. Other ontologiesare being developed to describe the structure of psychological tasks(CogPO: Turner and Laird (in press)) and EEG data (NEMO: http://nemo.nic.uoregon.edu/wiki/NEMO). The implementation of theseontologies will provide the semantic infrastructure that is needed tomaximize the power of shared data. In addition, thinking during thedesign process about how a specific task manipulation relates to anontology forces clarity about what the manipulation is measuring,which could be generally useful in driving cleaner and more specificexperimental designs.

Mathematical/computational models provide another languagefor describing the underlying processes being mapped by neuroim-

aging studies. In some areas of cognitive neuroscience (such as high-level vision and decision making), such models have already providedvaluable insights into the functional organization of the brain. Onedifficult but important challenge will be to effect a clearer mappingbetween the functional organization described by ontologies and thecomputational organization implemented in mathematical models.Within cognitive science, a number of general computationalframeworks have been developed for the description of cognitiveprocesses (e.g., Anderson, 1983; Chater et al., 2006; Rumelhart andMcClelland, 1986), and models based on these frameworks have beenrelated to brain function (e.g., Anderson et al., 2004; Norman andO'Reilly, 2003; Schultz et al., 1997); in the future, I expect thatcomputational frameworks like these will play an increasinglyimportant role in characterizing the mental operations being mappedusing neuroimaging.

Conclusions

fMRI has advanced cognitive neuroscience research in a way thathas been nothing short of revolutionary, though at the same timethere are fundamental limits to the standard imaging approach thathave not been widely appreciated. I am hopeful that 20 years fromnow, the history of fMRI in cognitive neuroscience will show that thefield attacked this problem head on and developed new, robustmethods for better understanding the relation between mentalprocesses and brain function.

Acknowledgments

Correspondence should be addressed to Russell A. Poldrack,University of Texas at Austin, 3925-B W. Braker Lane, Austin, TX78759 (email: poldrack@mail.utexas.edu). Thanks to Tyler Davis, KojiJimura, Jeanette Mumford, Tom Schonberg, Corey White, and TalYarkoni for helpful comments on a draft of this paper. Preparation ofthis manuscript was supported by NIH RO1MH082795.

References

Anderson, J.R., 1983. The Architecture of Cognition. Harvard University Press,Cambridge, MA.

Anderson, J.R., Bothell, D., Byrne, M.D., Douglass, S., Lebiere, C., Qin, Y., 2004. Anintegrated theory of the mind. Psychol. Rev. 111, 1036–1060.

Bard, J.B.L., Rhee, S.Y., 2004. Ontologies in biology: design, applications and futurechallenges. Nat. Rev. Genet. 5, 213–222.

Bennett, C.M., Wolford, G.L., Miller, M.B., 2009. The principled control of false positivesin neuroimaging. Soc. Cogn. Affect. Neurosci. 4, 417–422.

Bennett, C.M., Baird, A.A., Miller, M.B., Wolford, G.L., 2010. Neural correlates ofinterspecies perspective taking in the post-mortem Atlantic salmon: an argumentfor proper multiple comparisons correction. J. Serendipitous Unexpected Results 1,1–5.

Bilder, R.M., Sabb, F.W., Parker, D.S., Kalar, D., Chu, W.W., Fox, J., Freimer, N.B., Poldrack,R.A., 2009. Cognitive ontologies for neuropsychiatric phenomics research. Cogn.Neuropsychiatry 14, 419–450.

Biswal, B.B., Mennes, M., Zuo, X.N., Gohel, S., Kelly, C., Smith, S.M., Beckmann, C.F.,Adelstein, J.S., Buckner, R.L., Colcombe, S., Dogonowski, A.M., Ernst, M., Fair, D.,Hampson, M., Hoptman, M.J., Hyde, J.S., Kiviniemi, V.J., Kötter, R., Li, S.J., Lin, C.P.,Lowe, M.J., Mackay, C., Madden, D.J., Madsen, K.H., Margulies, D.S., Mayberg, H.S.,McMahon, K., Monk, C.S., Mostofsky, S.H., Nagel, B.J., Pekar, J.J., Peltier, S.J., Petersen,S.E., Riedl, V., Rombouts, S.A.R.B., Rypma, B., Schlaggar, B.L., Schmidt, S., Seidler, R.D.,Siegle, G.J., Sorg, C., Teng, G.J., Veijola, J., Villringer, A., Walter, M., Wang, L.,Weng, X.C., Whitfield-Gabrieli, S., Williamson, P., Windischberger, C., Zang, Y.F.,Zhang, H.Y., Castellanos, F.X., Milham, M.P., 2010. Toward discovery science ofhuman brain function. Proc. Natl. Acad. Sci. U.S.A. 107, 4734–4739.

Chater, N., Tenenbaum, J.B., Yuille, A., 2006. Probabilistic models of cognition:conceptual foundations. Trends Cogn. Sci. 10, 287–291.

David, O., Guillemain, I., Saillet, S., Reyt, S., Deransart, C., Segebarth, C., Depaulis, A.,2008. Identifying neural drivers with functional MRI: an electrophysiologicalvalidation. PLoS Biol. 6, 2683–2697.

Doan, A., Ramakrishnan, R., Halevy, A.Y., 2011. Crowdsourcing systems on the world-wide web. Commun. ACM 54, 86–96.

Huber, P.J., 2004. Robust Statistics. Wiley-Interscience, Hoboken, N.J.Ioannidis, J.P.A., 2005. Why most published research findings are false. PLoS Med. 2,

e124.Jimura, K., Poldrack, R.A., in press. Analyses of regional-average activation and

multivoxel pattern information tell complementary stories. Neuropsychologia.

1220 R.A. Poldrack / NeuroImage 62 (2012) 1216–1220

Kriegeskorte, N., Goebel, R., Bandettini, P., 2006. Information-based functional brainmapping. Proc. Natl. Acad. Sci. U.S.A. 103, 3863–3868.

Kriegeskorte, N., Mur, M., Bandettini, P., 2008. Representational similarity analysis —

connecting the branches of systems neuroscience. Front. Syst. Neurosci. 2, 4.Kriegeskorte, N., Simmons, W.K., Bellgowan, P.S.F., Baker, C.I., 2009. Circular analysis in

systems neuroscience - the dangers of double dipping. Nature Neuroscience 12 (5),535–540.

Laird, A.R., Lancaster, J.L., Fox, P.T., 2005. Brainmap: the social evolution of a humanbrain mapping database. Neuroinformatics 3, 65–78.

Norman, K.A., O'Reilly, R.C., 2003. Modeling hippocampal and neocortical contributionsto recognition memory: a complementary-learning-systems approach. Psychol.Rev. 110, 611–646.

Norman, K.A., Polyn, S.M., Detre, G.J., Haxby, J.V., 2006. Beyond mind-reading: multi-voxel pattern analysis of fMRI data. Trends Cogn. Sci. 10, 424–430.

Pereira, F., Mitchell, T., Botvinick, M., 2009. Machine learning classifiers and fMRI: atutorial overview. Neuroimage 45, S199–S209.

Poldrack, R.A., 2006. Can cognitive processes be inferred from neuroimaging data?Trends Cogn. Sci. 10, 59–63.

Poldrack, R.A., 2008. The role of fMRI in cognitive neuroscience: where do we stand?Curr. Opin. Neurobiol. 18, 223–227.

Poldrack, R.A., 2010. Mapping mental function to brain structure: how can cognitiveneuroimaging succeed? Perspect. Psychol. Sci. 5, 753–761.

Poldrack, R.A., Halchenko, Y.O., Hanson, S.J., 2009.Decoding the large-scale structure of brainfunction by classifying mental states across individuals. Psychol. Sci. 20, 1364–1372.

Poldrack, R.A., Mumford, J.A., Nichols, T.E., 2011. Handbook of fMRI Data Analysis.Cambridge University Press.

Poldrack, R.A., Kittur, A., Kalar, D., Miller, E., Seppa, C., Gil, Y., Parker, D.S., Sabb, F.W.,Bilder, R.M., 2011. The Cognitive Atlas: towards a knowledge foundation forcognitive neuroscience. Frontiers in Neuroinformatics 5, 17.

Ramsey, J.D., Hanson, S.J., Hanson, C., Halchenko, Y.O., Poldrack, R.A., Glymour, C., 2010.Six problems for causal inference from fMRI. Neuroimage 49, 1545–1558.

Rumelhart, D.E., McClelland, J.L., 1986. Parallel distributed processing: explorations inthe microstructure of cognition. MIT Press, Cambridge, Mass.

Schultz, W., Dayan, P., Montague, P.R., 1997. A neural substrate of prediction andreward. Science 275, 1593–1599.

Smith, S.M., Miller, K.L., Salimi-Khorshidi, G., Webster, M., Beckmann, C.F., Nichols, T.E.,Ramsey, J.D., Woolrich, M.W., 2011. Network modelling methods for fMRI.Neuroimage 54, 875–891.

Tom, S.M., Fox, C.R., Trepel, C., Poldrack, R.A., 2007. The neural basis of loss aversion indecision-making under risk. Science 315, 515–518.

Turner, J.A., Laird, A.R., in press. The cognitive paradigm ontology: Design andapplication. Neuroinformatics.

Van Horn, J.D., Grethe, J.S., Kostelec, P., Woodward, J.B., Aslam, J.A., Rus, D., Rockmore,D., Gazzaniga, M.S., 2001. The functional magnetic resonance imaging data center(fMRIDC): the challenges and rewards of large-scale databasing of neuroimagingstudies. Philos. Trans. R. Soc. Lond. B Biol. Sci. 356, 1323–1339.

Vul, E., Harris, C., Winkielman, P., Pashler, H., 2009. Puzzlingly high correlations in fMRIstudiesof emotion,personality, and social cognition. Perspect. Psychol. Sci. 4, 274–290.

Wager, T.D., Keller, M.C., Lacey, S.C., Jonides, J., 2005. Increased sensitivity inneuroimaging analyses using robust regression. Neuroimage 26, 99–113.

Woolrich, M., 2008. Robust group analysis using outlier inference. Neuroimage 41,286–301.

Xue, G., Dong, Q., Chen, C., Lu, Z., Mumford, J.A., Poldrack, R.A., 2010. Greater neural patternsimilarity across repetitions is associated with better memory. Science 330, 97–101.

Yarkoni, T., Poldrack, R.A., Van Essen, D.C., Wager, T.D., 2010. Cognitive neuroscience 2.0:building a cumulative science of humanbrain function. TrendsCogn. Sci. 14, 489–496.

Yarkoni, T., Poldrack, R.A.,Nichols, T.E.,VanEssen,D.,Wager, T., 2011. Large-scale automatedsynthesis of human functional neuroimaging data. Nat. Methods 8, 665–670.