Eyes-Open and Eyes-Closed Resting States With Opposite Brain Activity in Sensorimotor and Occipital Regions: Multidimensional Evidences From Machine Learning Perspective

Abstract

Studies have demonstrated that there are widespread significant differences in spontaneous brain activity between eyes-open (EO) and eyes-closed (EC) resting states. However, it remains largely unclear whether spontaneous brain activity is effectively related to EO and EC resting states. The amplitude, local functional concordance, inter-hemisphere functional synchronization, and network centrality of spontaneous brain activity were measured by the fraction amplitude of low frequency fluctuation (fALFF), regional homogeneity (ReHo), voxel-mirrored homotopic connectivity (VMHC) and degree centrality (DC), respectively. Using the public Eyes-open/Eyes-closed dataset, we employed the support vector machine (SVM) and bootstrap technique to establish linking models for the fALFF, ReHo, VMHC and DC dimensions. The classification accuracies of linking models are 0.72 (0.59, 0.82), 0.88 (0.79, 0.97), 0.82 (0.74, 0.91) and 0.70 (0.62, 0.79), respectively. Specifically, we observed that brain activity in the EO condition is significantly greater in attentional system areas, including the fusiform gyrus, occipital and parietal cortex, but significantly lower in sensorimotor system areas, including the precentral/postcentral gyrus, paracentral lobule (PCL) and temporal cortex compared to the EC condition from the four dimensions. The results consistently indicated that spontaneous brain activity is effectively related to EO and EC resting states, and the two resting states are of opposite brain activity in sensorimotor and occipital regions. It may provide new insight into the neural substrate of the resting state and help computational neuroscientists or neuropsychologists to choose an appropriate resting state condition to investigate various mental disorders from the resting state functional magnetic resonance imaging (fMRI) technique.

Decoding visual network‐related dynamic functional connectivity for eyes‐open and eyes‐closed using machine learning

“…2 Furthermore, our results found additional attention-related ICNs (ie, LFPN, RFPN, and DAN) that were altered between EO/EC, implying the different mechanisms of visual attention under the two resting states. 16 Moreover, the discriminative patterns were reproducible using two independent data sets. In summary, characteristic patterns of visual networkrelated dynamic FC were found between EO/EC, which provided a new perspective to investigate the brain mechanisms under EO/EC resting states.…”

“…Machine learning approaches have been applied to build discriminative models for EO/EC. 10,[14][15][16] The goals of discriminative models for EC/EC identification are twofold: to discover the neural mechanisms for EO/EC using data mining and to classify the states of EO/EC using machine learning. On the one hand, revealing the discriminative patterns under EO/EC using resting state fMRI could help better understanding the neural mechanisms of directional visual attention.…”

“…14,16 However, a feature selection procedure was required before support vector machine (SVM), since the dimensions of voxel-wise measures were relatively high. 14,16 In regional level, the sensorimotor regions-related fractional ALFF (fALFF) could classify EO/EC states using SVM. 15 However, the regional measures represented limited information within the region of interests (ROIs).…”

Resting state EEG as a biomarker of Parkinson’s disease: Influence of measurement conditions

“…Our data shows that electrodes over the motor cortex were valuable for the classifier, suggesting that activity in the motor cortex was a key contributor for the classification. Motor cortex activity is modulated by whether eyes are open or closed 7,25,32 , and PD, of course, is characterized by motor symptom. Electrodes located in the occipital pole also provided valuable information for the classifier, suggesting that visual cortical activity can be used to differentiate PD from healthy controls.…”

“…Our third aim was to examine if classification is influenced by whether the resting state EEG is recorded eyes open or closed. Because there are substantial differences in EEG dynamics between eyes closed and open conditions 2,12,13 -modulating, for instance, motor-related neural activity 7,25,32 -it could also affect classification performance. Most previous studies which classify PD based on resting state EEG have used eyes closed recordings 6,13,17 .…”