Differences

This shows you the differences between two versions of the page.

Link to this comparison view

Both sides previous revision Previous revision
tutorial:beamformingextended [2017/08/17 11:21]
127.0.0.1 external edit
tutorial:beamformingextended [2018/01/17 11:42] (current)
134.100.99.137
Line 20: Line 20:
 The brain is divided in a regular three dimensional grid and the source strength for each grid point is computed. The method applied in this example is termed Dynamical Imaging of Coherent Sources (DICS) and the estimates are calculated in the frequency domain (Gross et al. 2001). Other beamformer methods rely on sources estimates calculated in the time domain, e.g. the Linearly Constrained Minimum Variance (LCMV) and Synthetic Aperture Magnetometry (SAM) methods (van Veen et al., 1997; Robinson and Cheyne, 1997). These methods produce a 3D spatial distribution of the power of the neuronal sources. This distribution is then overlaid on a structural image of the subject'​s brain. These distributions of source power can then be subjected to statistical analysis. It is always ideal to contrast the activity of interest against some control/​baseline activity. Options for this will be discussed below, but it is best to keep this in mind when designing your experiment from the start, rather than struggle to find a suitable control/​baseline after data collection. The brain is divided in a regular three dimensional grid and the source strength for each grid point is computed. The method applied in this example is termed Dynamical Imaging of Coherent Sources (DICS) and the estimates are calculated in the frequency domain (Gross et al. 2001). Other beamformer methods rely on sources estimates calculated in the time domain, e.g. the Linearly Constrained Minimum Variance (LCMV) and Synthetic Aperture Magnetometry (SAM) methods (van Veen et al., 1997; Robinson and Cheyne, 1997). These methods produce a 3D spatial distribution of the power of the neuronal sources. This distribution is then overlaid on a structural image of the subject'​s brain. These distributions of source power can then be subjected to statistical analysis. It is always ideal to contrast the activity of interest against some control/​baseline activity. Options for this will be discussed below, but it is best to keep this in mind when designing your experiment from the start, rather than struggle to find a suitable control/​baseline after data collection.
  
-When conducting a multiple-subject study, it is essential that averaging over subjects does not violate any statistical assumption. One of these assumptions is that subject'​s sources are represented in a common space, i.e. an averaged grid point represents the estimate of the same brain region across subjects. One way to get subjects in a common space is by spatially deforming and interpolating the source reconstruction after beamforming. However, we will use an alternative way that does not require interpolation. Prior to source estimation we construct a regular grid in MNI template space and spatially deform this grid to each of the individual subjects (note that you will only have the data from one subject here). The beamformer estimation is done on the direct grid mapped to MNI space, so that the results can be compared over subjects. This procedure is explained in detail [[example/​create_single-subject_grids_in_individual_head_space_that_are_all_aligned_in_mni_space|in this example code]]. Creating the MNI template grid only needs to be done once, and the result ​in provided in the fieldtrip/​template directory. We strongly suggest that you have a quick (but thorough) look at the example code page and understand the essence of what is being done there anyway! ​+When conducting a multiple-subject study, it is essential that averaging over subjects does not violate any statistical assumption. One of these assumptions is that subject'​s sources are represented in a common space, i.e. an averaged grid point represents the estimate of the same brain region across subjects. One way to get subjects in a common space is by spatially deforming and interpolating the source reconstruction after beamforming. However, we will use an alternative way that does not require interpolation. Prior to source estimation we construct a regular grid in MNI template space and spatially deform this grid to each of the individual subjects (note that you will only have the data from one subject here). The beamformer estimation is done on the direct grid mapped to MNI space, so that the results can be compared over subjects. This procedure is explained in detail [[example/​create_single-subject_grids_in_individual_head_space_that_are_all_aligned_in_mni_space|in this example code]]. Creating the MNI template grid only needs to be done once, and the result ​is provided in the fieldtrip/​template directory. We strongly suggest that you have a quick (but thorough) look at the example code page and understand the essence of what is being done there anyway! ​
  
 The tutorial is split into three parts. In the first part of the tutorial, we will explain how to compute the forward and inverse model, which is the fundamental basic for source level analysis. In the second part, we will localize the sources responsible for the posterior gamma activity upon visual stimlation. In the third part of the tutorial, we will compute coherence to study the oscillatory synchrony between two sources in the brain. This is computed in the frequency domain by normalizing the magnitude of the summed cross-spectral density between two signals by their respective power. For each frequency bin the coherence value is a number between 0 and 1. The coherence values reflect the consistency of the phase difference between the two signals at a given frequency. In the dataset we will analyse the subject was required to maintain an isometric contraction of a forearm muscle. The example in this session covers thus cortico-muscular coherence on source level. The same principles, however, apply to cortico-cortical coherence, for which the interested reader can already have a look at [[tutorial:​connectivityextended| another tutorial]] that will be covered later. The tutorial is split into three parts. In the first part of the tutorial, we will explain how to compute the forward and inverse model, which is the fundamental basic for source level analysis. In the second part, we will localize the sources responsible for the posterior gamma activity upon visual stimlation. In the third part of the tutorial, we will compute coherence to study the oscillatory synchrony between two sources in the brain. This is computed in the frequency domain by normalizing the magnitude of the summed cross-spectral density between two signals by their respective power. For each frequency bin the coherence value is a number between 0 and 1. The coherence values reflect the consistency of the phase difference between the two signals at a given frequency. In the dataset we will analyse the subject was required to maintain an isometric contraction of a forearm muscle. The example in this session covers thus cortico-muscular coherence on source level. The same principles, however, apply to cortico-cortical coherence, for which the interested reader can already have a look at [[tutorial:​connectivityextended| another tutorial]] that will be covered later.