Limits...
Revealing instances of coordination among multiple cortical areas.

Abeles M - Biol Cybern (2013)

Bottom Line: Cognitive functions must involve interactions between several (perhaps many) cortical regions.In these current dipole traces, instances of brief activity undulations (BAUs) are automatically detected and used to reveal where and when cortical points interact.The article shows that these BAUs truly represent the reorganization of activity at the cortex and are strongly connected to behavior.

View Article: PubMed Central - PubMed

Affiliation: Gonda Brain Research Center, Bar-Ilan University, Max and Anna Webb Str., 52900 , Ramat Gan, Israel, abelesm@mail.biu.ac.il.

ABSTRACT
Cognitive functions must involve interactions between several (perhaps many) cortical regions. The instances of such interactions may not be tightly time locked to any external cue. Thus averaging over repeated trials of brain activity or its spectrograms may miss these instances. Here, coordinated activity among multiple cortical locations is revealed in ongoing activity with millisecond accuracy without the need for averaging over time or frequencies. This is based on reconstructions of the cortical current dipole amplitudes at multiple points from MEG recordings. In these current dipole traces, instances of brief activity undulations (BAUs) are automatically detected and used to reveal where and when cortical points interact. The article shows that these BAUs truly represent the reorganization of activity at the cortex and are strongly connected to behavior.

Show MeSH

Related in: MedlinePlus

Confusion matrices for randomly shuffled data. Data for all 7 tasks (3 relaxing and 4 mental and motor tasks) were cut into 686 disjoint segments, mixed, and reassigned to 7 tasks so that each mixed task contained exactly 14 segments from each of the original tasks. This was repeated 100 times to obtain 100 sets, each with 7 mixed tasks. Left—the confusion matrix for mixture number 1. Right—histogram for the sum over the diagonal for all 100 mixtures. The red line is the sum for the unmixed data
© Copyright Policy - OpenAccess
Related In: Results  -  Collection


getmorefigures.php?uid=PMC4228107&req=5

Fig8: Confusion matrices for randomly shuffled data. Data for all 7 tasks (3 relaxing and 4 mental and motor tasks) were cut into 686 disjoint segments, mixed, and reassigned to 7 tasks so that each mixed task contained exactly 14 segments from each of the original tasks. This was repeated 100 times to obtain 100 sets, each with 7 mixed tasks. Left—the confusion matrix for mixture number 1. Right—histogram for the sum over the diagonal for all 100 mixtures. The red line is the sum for the unmixed data

Mentions: When the confusion matrix was computed for data shuffled among all 7 tasks, the results appear random (Fig. 8-left). Random shuffling was done 100 times and the distribution of the sums over the diagonal of the confusion matrix was around 1, whereas for the true data, it was 4.3 (Fig. 8-right). Clearly, categorizing the task based on the density of BAUs over the hull was not driven by chance.Fig. 8


Revealing instances of coordination among multiple cortical areas.

Abeles M - Biol Cybern (2013)

Confusion matrices for randomly shuffled data. Data for all 7 tasks (3 relaxing and 4 mental and motor tasks) were cut into 686 disjoint segments, mixed, and reassigned to 7 tasks so that each mixed task contained exactly 14 segments from each of the original tasks. This was repeated 100 times to obtain 100 sets, each with 7 mixed tasks. Left—the confusion matrix for mixture number 1. Right—histogram for the sum over the diagonal for all 100 mixtures. The red line is the sum for the unmixed data
© Copyright Policy - OpenAccess
Related In: Results  -  Collection

Show All Figures
getmorefigures.php?uid=PMC4228107&req=5

Fig8: Confusion matrices for randomly shuffled data. Data for all 7 tasks (3 relaxing and 4 mental and motor tasks) were cut into 686 disjoint segments, mixed, and reassigned to 7 tasks so that each mixed task contained exactly 14 segments from each of the original tasks. This was repeated 100 times to obtain 100 sets, each with 7 mixed tasks. Left—the confusion matrix for mixture number 1. Right—histogram for the sum over the diagonal for all 100 mixtures. The red line is the sum for the unmixed data
Mentions: When the confusion matrix was computed for data shuffled among all 7 tasks, the results appear random (Fig. 8-left). Random shuffling was done 100 times and the distribution of the sums over the diagonal of the confusion matrix was around 1, whereas for the true data, it was 4.3 (Fig. 8-right). Clearly, categorizing the task based on the density of BAUs over the hull was not driven by chance.Fig. 8

Bottom Line: Cognitive functions must involve interactions between several (perhaps many) cortical regions.In these current dipole traces, instances of brief activity undulations (BAUs) are automatically detected and used to reveal where and when cortical points interact.The article shows that these BAUs truly represent the reorganization of activity at the cortex and are strongly connected to behavior.

View Article: PubMed Central - PubMed

Affiliation: Gonda Brain Research Center, Bar-Ilan University, Max and Anna Webb Str., 52900 , Ramat Gan, Israel, abelesm@mail.biu.ac.il.

ABSTRACT
Cognitive functions must involve interactions between several (perhaps many) cortical regions. The instances of such interactions may not be tightly time locked to any external cue. Thus averaging over repeated trials of brain activity or its spectrograms may miss these instances. Here, coordinated activity among multiple cortical locations is revealed in ongoing activity with millisecond accuracy without the need for averaging over time or frequencies. This is based on reconstructions of the cortical current dipole amplitudes at multiple points from MEG recordings. In these current dipole traces, instances of brief activity undulations (BAUs) are automatically detected and used to reveal where and when cortical points interact. The article shows that these BAUs truly represent the reorganization of activity at the cortex and are strongly connected to behavior.

Show MeSH
Related in: MedlinePlus