Localization bias and spatial resolution of adaptive and non-adaptive spatial filters for MEG source reconstruction

Sekihara, Kensuke; Sahani, Maneesh; Nagarajan, Srikantan S.

doi:10.1016/j.neuroimage.2004.11.051

Cited by 443 publications

(368 citation statements)

References 21 publications

Supporting

Mentioning

360

Contrasting

Unclassified

Order By: Relevance

“…Indeed, the MEG signal generated by a single source affects many sensors (field spread), and in turn, source maps reconstructed using the most typical inverse solvers (LCMVB or MNE) are fraught with long-range correlation (spatial leakage). The way this correlation vanishes with distance depends on the location and orientation of the source (Hari et al, 1988), on the functional parameter under consideration, on the source reconstruction method used (Sekihara et al, 2005;Wens, 2015), and on the SNR in adaptive approaches (Barnes and Hillebrand, 2003;Gross et al, 2001;Van Veen et al, 1997). Also, inconsistent errors across subjects in MEG-MRI coregistration further increase the smoothness of group-averaged MEG maps (Singh et al, 1997).…”

Section: Contrast Maps In Other Imaging Modalitiesmentioning

confidence: 99%

“…Source maps reconstructed with classical inverse solutions might be fraught with such location bias due to many factors (Sekihara et al, 2005). The magnitude of such bias may also be modulated by the SNR, which could lead to false positive detection of a location difference.…”

Section: Applicability Of the Proposed Location-comparison Proceduresmentioning

confidence: 99%

“…Therefore, for the sake of comparing the location between conditions, it is necessary to opt for a source reconstruction method whose location bias is minimally affected by potential differences between conditions. To that aim, the same inverse solution should be applied to both conditions, as is common practice in MEG studies , and LCMVB appears to be a method of choice since it has no location bias even in the presence of noise (Sekihara et al, 2005). This theoretical nicety is however not always verified in practice and a depth bias may persist and be modulated by the SNR.…”

Section: Applicability Of the Proposed Location-comparison Proceduresmentioning

confidence: 99%

See 2 more Smart Citations

Contrasting functional imaging parametric maps: The mislocation problem and alternative solutions

2018

View full text Add to dashboard Cite

A B S T R A C TIn the field of neuroimaging, researchers often resort to contrasting parametric maps to identify differences between conditions or populations. Unfortunately, contrast patterns mix effects related to amplitude and location differences and tend to peak away from sources of genuine brain activity to an extent that scales with the smoothness of the maps. Here, we illustrate this mislocation problem on source maps reconstructed from magnetoencephalographic recordings and propose a novel, dedicated location-comparison method. In realistic simulations, contrast mislocation was on average~10 mm when genuine sources were placed at the same location, and was still above 5 mm when sources were 20 mm apart. The dedicated location-comparison method achieved a sensitivity of~90% when inter-source distance was 12 mm. Its benefit is also illustrated on real brain-speech entrainment data. In conclusion, contrasts of parametric maps provide precarious information for source location. To specifically address the question of location difference, one should turn to dedicated methods as the one proposed here.

show abstract

Section: Contrast Maps In Other Imaging Modalitiesmentioning

confidence: 99%

Section: Applicability Of the Proposed Location-comparison Proceduresmentioning

confidence: 99%

Section: Applicability Of the Proposed Location-comparison Proceduresmentioning

confidence: 99%

See 1 more Smart Citation

Contrasting functional imaging parametric maps: The mislocation problem and alternative solutions

2018

View full text Add to dashboard Cite

show abstract

“…In this respect, beamforming approaches, which are more focused 434 than minimum norm estimates (Sekihara et al, 2005) might be eligible to localize these sources. 435…”

Section: Paradigm 421mentioning

confidence: 99%

Being an agent or an observer: Different spectral dynamics revealed by MEG

et al. 2014

View full text Add to dashboard Cite

14Several neuroimaging studies reported that a common set of regions are recruited during 15 action observation and execution and it has been proposed that the modulation of the µ rhythm, 16 in terms of oscillations in the alpha and beta bands might represent the electrophysiological 17 correlate of the underlying brain mechanisms. However, the specific functional role of these bands 18 within the µ rhythm is still unclear. Here, we used magnetoencephalography (MEG) to analyze the 19 spectral and temporal properties of the alpha and beta bands in healthy subjects during an action 20 observation and execution task. 21We associated the modulation of the alpha and beta power to a broad action observation 22 network comprising several parieto--frontal areas previously detected in fMRI studies. Of note, we 23 observed a dissociation between alpha and beta bands with a slow--down of beta oscillations 24 ACCEPTED FOR PUBLICATION -NEUROIMAGE2 compared to alpha during action observation. We hypothesize that this segregation is linked to a 25 different sequence of information processing and we interpret these modulations in terms of 26 internal models (forward and inverse). In fact, these processes showed opposite temporal 27 sequences of occurrence: anterior--posterior during action (both in alpha and beta bands) and 28 roughly posterior--anterior during observation (in the alpha band). The observed differentiation 29 between alpha and beta suggests that these two bands might pursue different functions in the 30 action observation and execution processes. 32Keywords: action observation and execution, alpha and beta rhythms, Event--Related 33Desynchronization (ERD), magnetoencephalography (MEG), internal models 34 35 36 INTRODUCTION 37Several studies report that our sensorimotor system is activated when we observe an 38 action performed by other people. This putative mirror--like activity in humans was found in the 39 precentral gyrus (vPM), the inferior frontal gyrus (IFG), the inferior parietal lobule (IPL), and 40 regions within the intraparietal sulcus (for a review see Rizzolatti and Sinigaglia, 2010). In addition 41 to this limited number of regions, neuroimaging studies observed a broader action observation 42 network (AON) which seems to be involved during action observation (OBS) and execution (EXE) 43 (Avenanti et al., 2012;Buccino et al., 2004;Gazzola and Keysers, 2009). A proposed theoretical 44 framework postulates that such AON implements forward and inverse internal models (Wolpert 45 and Ghahramani, 2000), which should be engaged during EXE and OBS. The internal models can 46 account both for how we link our actions to sensory consequences and how others' actions match 47 our own actions and sensations (Gazzola and Keysers, 2009;Iacoboni, 2005). It has been proposed 48 that the fronto--parietal AON has a predictive nature and according to this hypothesis, during 49 action observation the inverse and forward models are integrated to achieve a prediction of 50 ACCEPTED FOR PUBLICATION -NEUROIMA...

show abstract

“…Effects of emotion and memory on scene recognition 16 active sources without any a priori assumption on the number and position of the underlying 334 dipoles (for a mathematical validation of this localization technique, see Sekihara et al, 2005). 335 sLORETA solutions are computed within a three-shell spherical head model co-registered to the 336 MNI152 template (Mazziotta et al, 2001), restricted to the gray matter and the hippocampus.…”

Section: Source Localization Analysis 328mentioning

confidence: 99%

Multiple synergistic effects of emotion and memory on proactive processes leading to scene recognition

Schettino

Loeys

2013

NeuroImage

View full text Add to dashboard Cite

21Visual scene recognition is a proactive process through which contextual cues and top-down 22 expectations facilitate the extraction of invariant features. Whether the emotional content of the 23 scenes exerts a reliable influence on these processes or not, however, remains an open question. 24Here, topographic ERP mapping analysis and a distributed source localization method were used 25 to characterize the electrophysiological correlates of proactive processes leading to scene 26 recognition, as well as the potential modulation of these processes by memory and emotion. On 27 each trial, the content of a complex neutral or emotional scene was progressively revealed, and 28 participants were asked to decide whether this scene had previously been encountered or not 29(delayed match-to-sample task). Behavioral results showed earlier recognition for old compared 30to new scenes, as well as delayed recognition for emotional vs. neutral scenes. 31Electrophysiological results revealed that, ~400 ms following stimulus onset, activity in ventral 32 object-selective regions increased linearly as a function of accumulation of perceptual evidence 33 prior to recognition of old scenes. The emotional content of the scenes had an early influence in 34 these areas. By comparison, at the same latency, the processing of new scenes was mostly 35 achieved by dorsal and medial frontal brain areas, including the anterior cingulate cortex and the 36 insula. In the latter region, emotion biased recognition at later stages, likely corresponding to 37 decision making processes. These findings suggest that emotion can operate at distinct and 38 multiple levels during proactive processes leading to scene recognition, depending on the extent 39 of prior encounter with these scenes. 40

show abstract

Localization bias and spatial resolution of adaptive and non-adaptive spatial filters for MEG source reconstruction

Cited by 443 publications

References 21 publications

Contrasting functional imaging parametric maps: The mislocation problem and alternative solutions

Contrasting functional imaging parametric maps: The mislocation problem and alternative solutions

Being an agent or an observer: Different spectral dynamics revealed by MEG

Multiple synergistic effects of emotion and memory on proactive processes leading to scene recognition

Contact Info

Product

Resources

About