Article

Assessing the spatiotemporal evolution of neuronal activation with single-trial event-related potentials and functional MRI. Proc Natl Acad Sci U S A

Department of Biological and Medical Psychology, University of Bergen, 5009 Bergen, Norway. tom.eichele@
Proceedings of the National Academy of Sciences (Impact Factor: 9.67). 01/2006; 102(49):17798-803. DOI: 10.1073/pnas.0505508102
Source: PubMed

ABSTRACT

The brain acts as an integrated information processing system, which methods in cognitive neuroscience have so far depicted in a fragmented fashion. Here, we propose a simple and robust way to integrate functional MRI (fMRI) with single trial event-related potentials (ERP) to provide a more complete spatiotemporal characterization of evoked responses in the human brain. The idea behind the approach is to find brain regions whose fMRI responses can be predicted by paradigm-induced amplitude modulations of simultaneously acquired single trial ERPs. The method was used to study a variant of a two-stimulus auditory target detection (odd-ball) paradigm that manipulated predictability through alternations of stimulus sequences with random or regular target-to-target intervals. In addition to electrophysiologic and hemodynamic evoked responses to auditory targets per se, single-trial modulations were expressed during the latencies of the P2 (170-ms), N2 (200-ms), and P3 (320-ms) components and predicted spatially separated fMRI activation patterns. These spatiotemporal matches, i.e., the prediction of hemodynamic activation by time-variant information from single trial ERPs, permit inferences about regional responses using fMRI with the temporal resolution provided by electrophysiology.

Download full-text

Full-text

Available from: Tom Eichele
  • Source
    • "Given the orthogonal nature of the benefits and limitations of EEG and fMRI, it is a natural question to ask whether these measures can be combined in an effort to capitalize on the temporal and spatial resolutions provided by each modality. There currently exist several methods for " fusing " multimodal functional neuroimaging data together, such as partial least squares correlation (Lin et al., 2003; Martinez-Montes et al., 2004 ), independent component analysis (Beckmann and Smith, 2005; Liu & Calhoun, 2007; Calhoun et al., 2006; Eichele et al., 2009; Franco et al., 2008; Teipel et al., 2010; Xu et al., 2009; Calhoun et al., 2011;), structural equation modeling (Astolfi et al., 2004; Hamandi et al., 2008), multiple regression (De Martino et al., 2010; Eichele et al., 2005), and canonical correlation analysis (Correa et al., 2010a; Correa et al., 2010b; Correa et al., 2008). However, the focus of these methods is either source localization (for reviews, see Sui et al. (2012); Dähne et al. (2015)), or relating behavioral measures (e.g., response times) to brain data (for a review, see Krishnan et al. (2015) . "
    [Show abstract] [Hide abstract]
    ABSTRACT: The need to test a growing number of theories in cognitive science has led to increased interest in inferential methods that integrate multiple data modalities. In this manuscript, we show how a method for integrating three data modalities within a single framework provides (1) more detailed descriptions of cognitive processes and (2) more accurate predictions of unobserved data than less integrative methods. Specifically, we show how combining either EEG and fMRI with a behavioral model can perform substantially better than a behavioral-data-only model in both generative and predictive modeling analyses. We then show how a trivariate model - a model including EEG, fMRI, and behavioral data - outperforms bivariate models in both generative and predictive modeling analyses. Together, these results suggest that within an appropriate modeling framework, more data can be used to better constrain cognitive theory, and to generate more accurate predictions for behavioral and neural data.
    Full-text · Article · Dec 2015 · NeuroImage
  • Source
    • "Modern neuroscience benefits greatly from a multitude of imaging techniques that, individually, have helped to further our understanding of cognitive processing [1], [2] and improved clinical diagnostics [3], [4]. The combination of several imaging modalities originated in the context of epilepsy imaging [5]–[7] but has since then become an important asset in cognitive neuroscience. "
    [Show abstract] [Hide abstract]
    ABSTRACT: Multimodal data is ubiquitous in engineering, communication, robotics, vision or more generally speaking in industry and the sciences. All disciplines have developed their respective sets of analytic tools to fuse the information that is available in all measured modalities. In this paper we provide a review of classical as well as recent machine learning methods (specifically factor models) for fusing information from functional neuroimaging techniques such as LFP, EEG, MEG, fNIRS and fMRI. Early and late fusion scenarios are distinguished and appropriate factor models for the respective scenarios are presented along with example applications from selected multimodal neuroimaging studies. Further emphasis is given to the interpretability of the resulting model parameters, in particular by highlighting how factor models relate to physical models needed for source localization. The methods we discuss allow to extract information from neural data, which ultimately contributes to (a) better neuroscientific understanding, (b) enhance diagnostic performance and (c) discover neural signals of interest that correlate maximally with a given cognitive paradigm. While we clearly study the multimodal functional neuroimaging challenge, the discussed machine learning techniques have a wide applicability beyond, i.e. in general data fusion and may thus be informative to the general interested reader.
    Full-text · Article · Aug 2015 · Proceedings of the IEEE
  • Source
    • " - specified sum ( Hugdahl et al . , 2004 ) . Beliefs about others , Theory of Mind test ( panel #8 ) , infer the reason of an observed action ( Specht et al . , in preparation ) . Context updating , oddball detection task ( panel #9 ) , press a button whenever heard a tone with a deviating pitch in a stream of standard tones with the same pitch ( Eichele et al . , 2005 ) . All tasks were standard cognitive or neuropsychology tasks or tests that had been adapted to the MR scanner environment , with a slight modification of the Stroop task that also had a working memory component . All tasks had about the same duration and involved either a motor or a verbal response , with approximately the same freque"
    [Show abstract] [Hide abstract]
    ABSTRACT: In this paper we suggest the existence of a generalized task-related cortical network that is up-regulated whenever the task to be performed requires the allocation of generalized non-specific cognitive resources, independent of the specifics of the task to be performed. We have labeled this general purpose network, the extrinsic mode network (EMN) as complementary to the default mode network (DMN), such that the EMN is down-regulated during periods of task-absence, when the DMN is up-regulated, and vice versa. We conceptualize the EMN as a cortical network for extrinsic neuronal activity, similar to the DMN as being a cortical network for intrinsic neuronal activity. The EMN has essentially a fronto-temporo-parietal spatial distribution, including the inferior and middle frontal gyri, inferior parietal lobule, supplementary motor area, inferior temporal gyrus. We hypothesize that this network is always active regardless of the cognitive task being performed. We further suggest that failure of network up- and down-regulation dynamics may provide neuronal underpinnings for cognitive impairments seen in many mental disorders, such as, e.g., schizophrenia. We start by describing a common observation in functional imaging, the close overlap in fronto-parietal activations in healthy individuals to tasks that denote very different cognitive processes. We now suggest that this is because the brain utilizes the EMN network as a generalized response to tasks that exceeds a cognitive demand threshold and/or requires the processing of novel information. We further discuss how the EMN is related to the DMN, and how a network for extrinsic activity is related to a network for intrinsic activity. Finally, we discuss whether the EMN and DMN networks interact in a common single brain system, rather than being two separate and independent brain systems.
    Full-text · Article · Aug 2015 · Frontiers in Human Neuroscience
Show more