# Complex Pearson Correlation Coefficient for EEG Connectivity Analysis

^{1}

^{2}

^{*}

## Abstract

**:**

## 1. Introduction

## 2. Methods

#### 2.1. Complex Pearson Correlation Coefficient as a Measure of Undirected Connectivity

#### 2.2. Phase Locking Value PLV and Its Relation to CPCC

#### 2.3. Weighted Phase Lag Index wPLI and Its Relation to CPCC

#### 2.4. Connectivity Estimation Based on Phase Difference Histograms

## 3. Results

#### 3.1. Synthetic Signals from the MRC Brain Network Dynamics Unit (University of Oxford)

#### 3.2. Synthetic Signals Generated with the Kuramoto Model

#### 3.3. Real-Life Signals

- The raw brain activity data were imported into MATLAB using the EEGLAB toolbox;
- Electrode positions (also called channel locations) were defined in the software;
- The data were referenced to average;
- The data were filtered with a band pass filter limited to 0.5 and 45 Hz;
- Automatic spectral-based channel suppression (z = 5) was performed using the EEGLAB “pop rejchan” function;
- Artifacts were removed using the ICLabel plugin for EEGLAB (thresholds for removing components were less than or equal to 0.05 for brain activity and greater than or equal to 0.9 for artifacts);
- The data were re-referenced to average;
- Sub-bands of the EEG signal were extracted (delta 0.5–4 Hz, theta 4–8 Hz, alpha 8–13 Hz, low beta 13–18 Hz, high beta 18–30 Hz, gamma 35–45 Hz).

## 4. Discussion and Conclusions

## Author Contributions

## Funding

## Institutional Review Board Statement

## Informed Consent Statement

## Data Availability Statement

## Conflicts of Interest

## Abbreviations

absCPCC | absolute value of complex Pearson correlation coefficient |

CPCC | complex Pearson correlation coefficient |

DTI | diffusion tensor imaging |

EC | eyes closed |

EEG | electroencephalography |

EO | eyes open |

HT | Hilbert transform |

imCPCC | imaginary component of complex Pearson correlation coefficient |

MEG | magnetoencephalography |

MRI | magnetic resonance imaging |

PLV | phase locking value |

PLI | phase lag index |

wPLI | weighted phase lag index |

## References

- Fornito, A.; Zalesky, A.; Bullmore, E. Fundamentals of Brain Network Analysis, 1st ed.; Academic Press: London, UK, 2016; pp. 1–3. [Google Scholar]
- Sanei, S.; Chambers, J.A. EEG Signal Processing, 1st ed.; John Wiley & Sons: Hoboken, NJ, USA, 2016; pp. 1–8. [Google Scholar]
- Olejarczyk, E.; Marzetti, L.; Pizzella, V.; Zappasodi, F. Comparison of connectivity analyses for resting state EEG data. J. Neural Eng.
**2017**, 14, 036017. [Google Scholar] [CrossRef] [PubMed] - Koch, M.A.; Norris, D.G.; Hund-Georgiadis, M. An investigation of functional and anatomical connectivity using magnetic resonance imaging. Neuroimage
**2002**, 16, 241–250. [Google Scholar] [CrossRef] [PubMed] [Green Version] - Rose, S.; Pannek, K.; Bell, C.; Baumann, F.; Hutchinson, N.; Coulthard, A.; McCombe, P.; Henderson, R. Direct evidence of intra- and interhemispheric corticomotor network degeneration in amyotrophic lateral sclerosis: An automated MRI structural connectivity study. Neuroimage
**2012**, 59, 2661–2669. [Google Scholar] [CrossRef] [PubMed] - Lim, S.; Han, C.E.; Uhlhaas, P.J.; Kaiser, M. Preferential detachment during human brain development: Age-and sex-specific structural connectivity in diffusion tensor imaging (DTI) data. Cereb. Cortex
**2015**, 25, 1477–1489. [Google Scholar] [CrossRef] [Green Version] - Singer, W. Neuronal synchrony: A versatile code for the definition of relations? Neuron
**1999**, 24, 49–65. [Google Scholar] [CrossRef] [Green Version] - Varela, F.; Lachaux, J.P.; Rodriguez, E.; Martinerie, J. The brainweb: Phase synchronization and large-scale integration. Nat. Rev. Neurosci.
**2001**, 2, 229–239. [Google Scholar] [CrossRef] - Fries, P. A mechanism for cognitive dynamics: Neuronal communication through neuronal coherence. Trends Cogn. Sci.
**2005**, 9, 474–480. [Google Scholar] [CrossRef] - Fries, P. Rhythms for cognition: Communication through coherence. Neuron
**2015**, 88, 220–235. [Google Scholar] [CrossRef] [Green Version] - Siegel, M.; Donner, T.H.; Engel, A.K. Spectral fingerprints of large-scale neuronal interactions. Nat. Rev. Neurosci.
**2012**, 13, 121–134. [Google Scholar] [CrossRef] - Bastos, A.M.; Schoffelen, J.M. A tutorial review of functional connectivity analysis methods and their interpretational pitfalls. Front. Syst. Neurosci.
**2016**, 9, 175. [Google Scholar] [CrossRef] [Green Version] - Sakkalis, V. Review of advanced techniques for the estimation of brain connectivity measured with EEG/MEG. Comput. Biol. Med.
**2011**, 41, 1110–1117. [Google Scholar] [CrossRef] - Hwa, R.C.; Ferree, T.C. Scaling properties of fluctuations in the human electroencephalogram. Phys. Rev. E
**2002**, 66, 021901. [Google Scholar] [CrossRef] [Green Version] - Li, L.; Zhang, J.X.; Jiang, T. Visual working memory load-related changes in neural activity and functional connectivity. PLoS ONE
**2011**, 6, e22357. [Google Scholar] [CrossRef] [Green Version] - Stam, C.J.; Van Dijk, B.W. Synchronization likelihood: An unbiased measure of generalized synchronization in multivariate data sets. Phys. D
**2002**, 163, 236–251. [Google Scholar] [CrossRef] - Stam, C.J.; Nolte, G.; Daffertshofer, A. Phase lag index: Assessment of functional connectivity from multi channel EEG and MEG with diminished bias from common sources. Hum. Brain Mapp.
**2007**, 28, 1178–1193. [Google Scholar] [CrossRef] - Lachaux, J.P.; Rodriguez, E.; Martinerie, J.; Varela, F.J. Measuring phase synchrony in brain signals. Hum. Brain Mapp.
**1999**, 8, 194–208. [Google Scholar] [CrossRef] [Green Version] - Vinck, M.; Oostenveld, R.; Van Wingerden, M.; Battaglia, F.; Pennartz, C.M. An improved index of phase-synchronization for electrophysiological data in the presence of volume-conduction, noise and sample-size bias. Neuroimage
**2011**, 55, 1548–1565. [Google Scholar] [CrossRef] - Schreier, P.J. A unifying discussion of correlation analysis for complex random vectors. IEEE Trans. Signal Process
**2008**, 56, 1327–1336. [Google Scholar] [CrossRef] - Ma, G.; Li, L. Depth and structural index estimation of 2D magnetic source using correlation coefficient of analytic signal. J. Appl. Geophys.
**2013**, 91, 9–13. [Google Scholar] [CrossRef] - Fuhrmann, D.R.; San Antonio, G. Transmit beamforming for MIMO radar systems using signal cross-correlation. IEEE Trans. Aerosp. Electron. Syst.
**2008**, 44, 171–186. [Google Scholar] [CrossRef] - Makita, S.; Kurokawa, K.; Hong, Y.J.; Miura, M.; Yasuno, Y. Noise-immune complex correlation for optical coherence angiography based on standard and Jones matrix optical coherence tomography. Biomed. Opt. Express
**2016**, 7, 1525–1548. [Google Scholar] [CrossRef] [Green Version] - Welch, P. The use of fast Fourier transform for the estimation of power spectra: A method based on time averaging over short, modified periodograms. IEEE Trans. Audio Electroacoust.
**1967**, 15, 70–73. [Google Scholar] [CrossRef] [Green Version] - Tass, P.; Rosenblum, M.G.; Weule, J.; Kurths, J.; Pikovsky, A.; Volkmann, J.; Schnitzler, A.; Freund, H.J. Detection of n:m Phase Locking from Noisy Data: Application to Magnetoencephalography. Phys. Rev. Lett.
**1998**, 81, 3291–3294. [Google Scholar] [CrossRef] - Mäkinen, V.; Tiitinen, H.; May, P. Auditory event-related responses are generated independently of ongoing brain activity. NeuroImage
**2005**, 24, 961–968. [Google Scholar] [CrossRef] - Šverko, Z.; Sajovic, J.; Drevenšek, G.; Vlahinić, S.; Rogelj, P. Generation of Oscillatory Synthetic Signal Simulating Brain Network Dynamics. In Proceedings of the 2021 44th International Convention on Information, Communication and Electronic Technology (MIPRO), MEET—Microelectronics, Electronics and Electronic Technology, Opatija, Croatia, 27 September–1 October 2021. [Google Scholar]
- Torkamani Azar, M.; Kanik, S.D.; Aydin, S.; Cetin, M. Prediction of reaction time and vigilance variability from spatio-spectral features of resting-state EEG in a long sustained attention task. IEEE J. Biomed. Health Inform.
**2020**, 24, 2550–2558. [Google Scholar] [CrossRef] [Green Version] - Fraschini, M.; Pani, S.M.; Didaci, L.; Marcialis, G.L. Robustness of functional connectivity metrics for EEG-based personal identification over task-induced intra-class and inter-class variations. Pattern Recognit. Lett.
**2019**, 125, 49–54. [Google Scholar] [CrossRef] - Hardmeier, M.; Hatz, F.; Bousleiman, H.; Schindler, C.; Stam, C.J.; Fuhr, P. Reproducibility of functional connectivity and graph measures based on the phase lag index (PLI) and weighted phase lag index (wPLI) derived from high resolution EEG. PLoS ONE
**2014**, 9, e108648. [Google Scholar] [CrossRef] [Green Version] - Hu, L.; Zhang, Z. EEG Signal Processing and Feature Extraction; Springer: Singapore, 2019; pp. 241–266. [Google Scholar]
- Sarmukadam, K.; Bitsika, V.; Sharpley, C.F.; McMillan, M.M.; Agnew, L.L. Comparing different EEG connectivity methods in young males with ASD. Behav. Brain Res.
**2020**, 383, 112482. [Google Scholar] [CrossRef] - Huang, H.; Zhang, J.; Zhu, L.; Tang, J.; Lin, G.; Kong, W.; Lei, X.; Zhu, L. EEG-Based Sleep Staging Analysis with Functional Connectivity. Sensors
**2021**, 21, 1988. [Google Scholar] [CrossRef] - Caicedo-Acosta, J.; Castaño, G.A.; Acosta-Medina, C.; Alvarez-Meza, A.; Castellanos-Dominguez, G. Deep Neural Regression Prediction of Motor Imagery Skills Using EEG Functional Connectivity Indicators. Sensors
**2021**, 21, 1932. [Google Scholar] [CrossRef] - Hag, A.; Handayani, D.; Pillai, T.; Mantoro, T.; Kit, M.H.; Al-Shargie, F. EEG Mental Stress Assessment Using Hybrid Multi-Domain Feature Sets of Functional Connectivity Network and Time-Frequency Features. Sensors
**2021**, 21, 6300. [Google Scholar] [CrossRef] [PubMed] - Katmah, R.; Al-Shargie, F.; Tariq, U.; Babiloni, F.; Al-Mughairbi, F.; Al-Nashash, H. A Review on Mental Stress Assessment Methods Using EEG Signals. Sensors
**2021**, 21, 5043. [Google Scholar] [CrossRef] - Formoso, M.A.; Ortiz, A.; Martinez-Murcia, F.J.; Gallego, N.; Luque, J.L. Detecting Phase-Synchrony Connectivity Anomalies in EEG Signals. Application to Dyslexia Diagnosis. Sensors
**2021**, 21, 7061. [Google Scholar] [CrossRef] [PubMed] - Anzolin, A.; Toppi, J.; Petti, M.; Cincotti, F.; Astolfi, L. SEED-G: Simulated EEG Data Generator for Testing Connectivity Algorithms. Sensors
**2021**, 21, 3632. [Google Scholar] [CrossRef] [PubMed] - Chikara, R.K.; Ko, L.-W. Modulation of the Visual to Auditory Human Inhibitory Brain Network: An EEG Dipole Source Localization Study. Brain Sci.
**2019**, 9, 216. [Google Scholar] [CrossRef] [PubMed] [Green Version] - Perez-Ortiz, C.X.; Gordillo, J.L.; Mendoza-Montoya, O.; Antelis, J.M.; Caraza, R.; Martinez, H.R. Functional Connectivity and Frequency Power Alterations during P300 Task as a Result of Amyotrophic Lateral Sclerosis. Sensors
**2021**, 21, 6801. [Google Scholar] [CrossRef] [PubMed] - García-Murillo, D.G.; Alvarez-Meza, A.; Castellanos-Dominguez, G. Single-Trial Kernel-Based Functional Connectivity for Enhanced Feature Extraction in Motor-Related Tasks. Sensors
**2021**, 21, 2750. [Google Scholar] [CrossRef] - Vecchio, F.; Pappalettera, C.; Miraglia, F.; Alù, F.; Orticoni, A.; Judica, E.; Cotelli, M.; Pistoia, F.; Rossini, P.M. Graph Theory on Brain Cortical Sources in Parkinson’s Disease: The Analysis of ‘Small World’ Organization from EEG. Sensors
**2021**, 21, 7266. [Google Scholar] [CrossRef]

**Figure 1.**Visualization of averaging used in calculation of the PLV. The PLV is computed from unit vectors representing instantaneous phase differences.

**Figure 2.**Visualization of averaging used in calculation of the PLI. The PLI is computed from unit vectors representing instantaneous phase differences.

**Figure 3.**The relationship between connectivity and phase differences distributions. When volume conduction is not considered (

**a**) higher connectivity reflects in lower variance, while the mean value is irrelevant. In the presence of volume conduction (

**b**), it reflects in higher values for phase differences close to 0 or $\pi $, which, therefore, do not (necessarily) indicate connectivity. Connectivity level is expressed with colors; red is the highest and yellow is the lowest.

**Figure 4.**Connectivity matrices obtained with PLV (

**a**), wPLI (

**b**), absCPCC (

**c**), and imCPCC (

**d**) for signals generated with [26], for 8–13 Hz frequency band.

**Figure 5.**Scatter plots of absCPCC to PLV relationship (

**left**) and imCPCC to wPLI relationship (

**right**). Each dot represents one electrode pair. Dots are colored according to their relative density. The black line represents identity, while the cyan one is the best linear fit. Rows correspond to different frequency bands: (

**a**,

**b**) 0.5–4 Hz; (

**c**,

**d**) 4–8 Hz; (

**e**,

**f**) 8–13 Hz.

**Figure 6.**Phase difference distributions for selected electrode pairs (synthetic signals [26]). Shown are the distributions corresponding to the highest: (

**a**) PLV and absCPCC values, (

**b**) wPLI and imCPCC values, (

**c**) ratio between absCPCC and imCPCC values, (

**d**) ratio between imCPCC and absCPCC values.

**Figure 7.**Connectivity matrices obtained with PLV, wPLI, absCPCC, and imCPCC for signals generated with the Kuramoto model [27].

**Figure 8.**Relationships between absCPCC and PLV (

**a**), and wPLI and imCPCC (

**b**), shown as a scatter plot of values for all signal pairs where signals were generated with the Kuramoto model [27]. The black line represent the identity while the cyan line shows the best linear fit. The colors of the dots represent the relative density of the connectivity values.

**Figure 9.**Phase difference distributions for selected synthetic signal pairs generated using the Kuramoto model [27]. Shown are the distributions corresponding to the highest: (

**a**) PLV and absCPCC values, (

**b**) wPLI and imCPCC values, (

**c**) ratio between absCPCC and imCPCC values, (

**d**) ratio between imCPCC and absCPCC values.

**Figure 10.**Connectivity matrices for the the alpha band (8–13 Hz) of the real-life signals [28] for eyes closed (EC) and eyes open (EO) states, computed with PLV, wPLI, absCPCC (

**g**), and imCPCC (

**h**).

**Figure 11.**Scatter plots of the absCPCC to PLV relationship (

**left**) and the imCPCC to wPLI relationship (

**right**), for all electrode pairs and for 10 test subjects (EC state). The black line represents the identity, while the cyan line shows the best linear fit. The colors of the dots represent the relative density of the connectivity values. Each row is shown for a different frequency band: (

**a**,

**b**) 0.5–4 Hz; (

**c**,

**d**) 4–8 Hz; (

**e**,

**f**) 8–13 Hz.

**Figure 12.**Scatter plots of the absCPCC to PLV relationship (

**left**) and the imCPCC to wPLI relationship (

**right**), for all electrode pairs and for 10 test subjects (EO state). The black line represents the identity, while the cyan line shows the best linear fit. The colors of the dots represent the relative density of the connectivity values. Each row is shown for a different frequency band: (

**a**,

**b**) 0.5–4 Hz; (

**c**,

**d**) 4–8 Hz; (

**e**,

**f**) 8–13 Hz.

**Figure 13.**Phase difference distributions for selected electrode pairs (real-life signals [28]). For each distribution, all four measures are calculated. The figure shows the electrode pair with the highest: (

**a**) PLV and absCPCC values, (

**b**) wPLI, and imCPCC values, (

**c**) ratio between absCPCC and imCPCC values, (

**d**) ratio between imCPCC and absCPCC values.

**Table 1.**Correlation values between compared connectivity measures (real-life signal). Here, ${r}_{abs}$ and ${r}_{im}$ denote $r(absCPCC,PLV)$ and $r(imCPCC,wPLI)$ respectively.

Frequency | State-EC | State-EO | ||
---|---|---|---|---|

(Hz) | ${\mathit{r}}_{\mathit{abs}}$ | ${\mathit{r}}_{\mathit{im}}$ | ${\mathit{r}}_{\mathit{abs}}$ | ${\mathit{r}}_{\mathit{im}}$ |

0.5–4 | 0.93 | 0.86 | 0.94 | 0.89 |

4–8 | 0.98 | 0.91 | 0.97 | 0.94 |

8–13 | 0.98 | 0.86 | 0.97 | 0.91 |

13–18 | 0.99 | 0.94 | 0.99 | 0.96 |

18–30 | 0.98 | 0.96 | 0.99 | 0.96 |

35–45 | 0.96 | 0.95 | 0.99 | 0.95 |

Publisher’s Note: MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations. |

© 2022 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https://creativecommons.org/licenses/by/4.0/).

## Share and Cite

**MDPI and ACS Style**

Šverko, Z.; Vrankić, M.; Vlahinić, S.; Rogelj, P.
Complex Pearson Correlation Coefficient for EEG Connectivity Analysis. *Sensors* **2022**, *22*, 1477.
https://doi.org/10.3390/s22041477

**AMA Style**

Šverko Z, Vrankić M, Vlahinić S, Rogelj P.
Complex Pearson Correlation Coefficient for EEG Connectivity Analysis. *Sensors*. 2022; 22(4):1477.
https://doi.org/10.3390/s22041477

**Chicago/Turabian Style**

Šverko, Zoran, Miroslav Vrankić, Saša Vlahinić, and Peter Rogelj.
2022. "Complex Pearson Correlation Coefficient for EEG Connectivity Analysis" *Sensors* 22, no. 4: 1477.
https://doi.org/10.3390/s22041477