03h+Coherence,+Phase+shift,+phase+lack,+effective+connectivity

**Any distortion** **of coherence is a distortion of identity** 





(Immune system??)

Old Coherence:

 * **Interaction size** under global setting: 1024 = 5sec of data at a sampling rate of 200/sec. The time is rather high. It could be better to make it shorter.
 * **Interaction part**: how many times shall I calculate the coherence value? It is set a 4/sec
 * **Smoothing type**: Frequency (was standard way to calculate it) and Time
 * **Coherence weighting**: takes into account the power of the specific frequency.

Coherence band

 * Average of all frequencies included. Allows weighing of the different frequencies.
 * Graphs:
 * Coherence:
 * Synchrony: phase relation is integrated. Synchrony you ask the signal exactly the same phase. It is more restricted than coherence
 * Similarity: these values are dependent of amplitude. That means that signals must have a high amplitude together with a high coherence.
 * Comodulation: the changes in Amplitude determine the value.
 * Feedback on Coherence, Synchrony, Similarity, Comodulation
 * Feedback levels
 * <span style="color: #0000ff; font-family: Papyrus,cursive; margin-bottom: 0cm;">Bandwidth: don’t make it too broad.
 * <span style="color: #0000ff; font-family: Papyrus,cursive; margin-bottom: 0cm;">Coherence time interval: Smoothing: the longer the time interval the lower the coherence.
 * <span style="color: #0000ff; font-family: Papyrus,cursive; margin-bottom: 0cm;">Cycles: how many cycles are included into calculation? 1 is used. That is not precise. That allows that neighbouring frequencies are included.

= **Coherences:** =


 * || **Result of decreased coherence** || **Result of increased coherence** ||
 * Fp1Fp2 || Less efficient integration of logical/emotional attention || Lack of flexibility in integrating logical/emotional attention ||
 * F3F4 || Less efficient motor actions RUE/motor actions LUE || Lack of flexibility motor actions RUE/motor actions LUE ||
 * F7F8 || Less efficient verbal/emotional expression || Lack of flexibility of verbal/emotional expression ||
 * C3C4 || Less efficient sensorimotor integration RUE / sensorimotor L || Lack of flexibility of sensorimotor integration RUE / sensorimotor L ||
 * T3T4 || Less efficient logical/emotional memory || Lack of flexibility of logical memory/emotional memory ||
 * P3P4 || Less efficient perceptions R / perceptions L || Lack of flexibility perceptions R / perceptions L ||
 * FzPz || Less efficient midline motor action / midline perception || Lack of flexibility of midline motor action / midline perception ||



COHERENCE, PHASE DIFFERENCES, PHASE SHIFT AND PHASE LOCK IN EEG/ERP ANALYSEStoc

Robert W. Thatcher, Ph.D.

[|Coherence, phase shift]

A review of the literature demonstrates that EEG coherence is related to a mixture of phase locking interrupted by phase shifts in the spontaneous EEG. EEG coherence is positively related to phase lock durations that range from approximately 100 msec to 400 msec in normal controls. Phase lock periods are interrupted for 30 to 80 msec by phase shifts which mark the beginning of a different set of connections that become phase locked and the process repeats continuously. Coherence is like a correlation coefficient and lacks temporal resolution in a time series and is influenced by volume conduction when phase difference = 0. Average reference and Laplacian transforms distort physiologically generated phase differences and invalidate the use of these transforms in the computation of coherence and phase differences as well as in the computation of directed coherence and phase reset. Time domain measures of phase shift and phase lock are less prone to artifact and are

independent of volume conduction. Another advantage of measuring network coupling in the time domain is that basic neurophysiological processes such as IPSPs and EPSP durations can be directly related to the dynamics measured in ERP and EEG studies.

=Effective connectivity (phase slope index)= uses estimates of the direction and magnitude of information transfer in a network {1, p.49}

Structural connectivity revealed by MRI nod DTI reveal the existence of a street, EEG coherence estimates the extent to which the street is used and effective connectivity estimates the number of people and the direction.