key: cord-0308094-bw92r5v9 authors: Lahijanian, Mojtaba; Aghajan, Hamid; Vahabi, Zahra; Afzal, Arshia title: Gamma Entrainment Improves Synchronization Deficits in Dementia Patients date: 2021-10-10 journal: bioRxiv DOI: 10.1101/2021.09.30.462389 sha: 337ad3c14f18b80529045056a0253e73eb174c9b doc_id: 308094 cord_uid: bw92r5v9 Non-invasive gamma entrainment has shown promising results in alleviating cognitive symptoms of Alzheimer’s disease in mice and humans. In this study, we examine improvements in the synchronization characteristics of the brain’s oscillations induced by 40Hz auditory stimulation based on electroencephalography data recorded from a group of dementia patients. We observed that when the quality of entrainment surpasses a certain level, several indicators of brain synchronization significantly improve. Specifically, the entrained oscillatory activity maintains temporal phase stability in the frontal, parietal, and occipital regions, and persistent spatial phase coupling between them. In addition, notable theta-gamma phase-amplitude coupling is observed in these areas. Interestingly, a high theta power at rest predicts the quality of entrainment. We identify differentiating attributes of temporal/spatial synchronization and cross-frequency coupling in the data of two groups with entrained and non-entrained responses which point to enhanced network synchronization caused by entrainment and can explain its potential therapeutic effects. Occurrence of entrainment during stimulation cycles. (a) Checkered diagrams show the occurrence of entrainment for all trials (columns), all channels (rows), and all participants in the entrained group (n = 4). The color of each matrix element represents a z-score value of the 40Hz component's amplitude in the frequency response and dark blue elements indicate entrained responses (z-score above 3). The star next to each channel implies that the majority of trial responses are entrained. (b) Same as a for the non-entrained group (n = 7). (c) Topographic distribution of entrainment over the scalp for the two groups. The gamma entrainment score (trial-averaged z-score, see methods for more information) is color coded and the plots show the average group response to the external stimulus. The frontal, parietal, and occipital channels show highest entrained responses in the entrained group. Ent.: Entrained, Non-ent.: Non-entrained. Intrinsic theta power is highly correlated with entrainability of gamma oscillations. To 123 examine the relationship between the power of theta and gamma oscillations on channel Fz, 124 we calculate ensemble averages for the stimulation and rest trials separately. We normalize the 125 power of the target frequency band during stimulation with the power measured during the rest 126 trials. 127 As Fig. 2a illustrates, the normalized gamma power response on Fz is considerably higher for 128 the entrained group compared to the non-entrained group. This difference is significant (ttest2, 129 t9 = 3.0111, p-value = 0.0147; Fig. 2c right) and does also exist between the two groups across 130 most other channels (as evidenced in Fig. 1c ). Fig. 2b indicates that the normalized theta power While normalized power can be used to determine the stimulation's effect compared to rest, it 134 cannot help assess the absolute power-differences between the two groups for the stimulation 135 and rest cycles separately. Thus, we further analyze power spectra without normalization. Fig. 136 2d shows that the theta power is significantly higher for the entrained group in both stimulation 137 (ttest2, t9 = 5.3826, p-value = 4.43e-04) and rest (ttest2, t9 = 6.6534, p-value = 9.34e-05) 138 intervals. In other words, high theta power is present during the entire task in the entrained 139 group. The concurrent presence of high theta power when gamma entrainment occurs can serve 140 as a possible explanation for the superior performance of the entrained group in terms of the 141 characteristics of the involved oscillatory bands. Remarkably, the presence of high theta power 142 in the EEG signal recorded one minute before the start of the task can indeed serve as a 143 predictor of the quality of the ensuing gamma entrainment (see Fig. 2e ; ttest2, t8 = 4.3778, p-144 value = 0.0024). To further examine the role of theta power in facilitating gamma entrainment, 145 we evaluate the correlation between them. As Fig. 2f indicates, not only higher theta power is 146 present in the entrained group, it also registers higher values with higher entrainment scores. 147 The relatively large positive correlation coefficient between the gamma entrainment score and Gamma power during stimulus cycles in 1sec sliding windows with 0.5sec overlap normalized to the total power of gamma in rest cycles for the entrained and non-entrained groups. (b) Same as a for theta power. (c) Total normalized (stimulus over rest) theta power (left) and gamma power (right) for both groups. (d) Theta power for stimulus (left) and rest (right) cycles for both groups. (e) Rest-state theta power for both groups calculated over 60sec before the start of the main task (rest-state data was not recorded from participant S9 in the non-entrained group). (f) Theta power versus gamma entrainment score. The entrained response is highly correlated with the theta power during stimulus cycles in the entrained group. This correlation is very low for the non-entrained group (Pearson correlation is used). (g) Same as d for the gamma power. Data expressed as mean ± SEM in shaded format in a-b and black lines in c-e, g as well as box-and-whisker plots in gray, violin plots representing the estimated normal distribution for each group, empty circles corresponding to each participant and white circles showing the median in the violin plots. *, p < 0.05; ***, p < 0.001. n.s.: not significant, Ent.: Entrained, Non-ent.: Non-entrained. area of the brain. Moreover, it is noteworthy that activity with persistent phase occurs at all Figure 3 Neural activity is highly synchronized during stimulation in the entrained group. (a) Polar histograms of unit-amplitude phasors of the 40Hz component in 1sec sliding windows for the stimulus and rest cycles for a representative participant (S1) in the entrained group, recorded from the frontal (top) and parietal/occipital (bottom) channels. The colored circles on the perimeter show the phases of each 1sec sample. The empty circles represent average phasors of each 1sec record (for stimulus and rest sets). The black arrows correspond to the mean vector for the stimulus cycle set, with large values indicating highly concentrated response phases. Similar mean angles are observed for the frontal (F3, Fz, F4) channels. Also similar mean angles are observed for the parietal/occipital (P3, Pz, P4, O1, O2) channels with an approximate 180-degree disparity to the frontal mean angles. All these phenomena are observed in all participants in the entrained group (see supplementary The entrained brain oscillations are spatially coupled. As discussed above and illustrated in Finally, the induced gamma oscillations are coupled with the phase of theta activity as 296 evidenced by high PAC values in comodulograms like Fig. 5a . This coupling follows a form 297 of inter-band synchronization reported to occur in binding activities that involve the hippocampus 13,44,48 . The similarity of the spatial topographies of high gamma power (Fig. 1c) 299 and high theta-gamma PAC (Fig. 5c) indicates that the entrained gamma oscillations tend to 300 couple their amplitudes with the phase of the underlying theta activity present in the network. How can the entrainment of gamma band oscillatory activity lead to the therapeutic effects 302 reported for AD 27-30,33,34 ? One mechanism which may help explain these effects is the principle 303 of neuroplasticity, which states that the synaptic links between neurons experiencing 304 synchronized activity strengthen over time. Through enticing simultaneous neural activity, 305 entrainment contributes to binding neural pathways and boosting of synaptic weights in 306 neuronal populations that are forced to undergo synchronized activity. The role of theta oscillations in entrainment. There still remain other important questions. What is the role of theta oscillations in the process of entrainment when the stimulant frequency 309 lies in the gamma band? Theta power plays a key role in human memory 48 . One observation in 310 our study is that theta oscillations matter in entrainment as high theta power leads to more 311 intensive entrainment (Fig. 2f) . To justify this effect, we can consider the process of theta- These points aside, the role of the theta-band activity should not be overlooked in entrainment 340 studies as there are evidences on improving performance in memory tasks due to theta 341 entrainment 53 . Our study led to remarkable observations on the role of theta power in gamma 342 entrainment, namely its very high correlation with the gamma entrainment score (Fig. 2f) , and the efficacy of the rest-state theta power in predicting successful entrainment (Fig. 2e) . These 344 observations elude to the idea that the presence of a theta oscillatory component in the stimulant 345 might improve the quality of gamma entrainment in non-entrained participants through 346 increasing theta activity in the brain. Synthetic auditory stimulants can be created in different 347 ways to contain both gamma and theta oscillations. In a recent report 54 , we used chirp segments 348 of natural canary song, which contains some level of intrinsic theta-gamma coupling, and 349 observed high-quality brain entrainment for a group of participants. Table 1 . 409 General exclusion criteria were: a history of stroke, traumatic brain injury, schizophrenia, (1) For the distance of a distribution ( ) from a uniform distribution , this equation simplifies where N is the number of bins. and Pz channels, respectively, the PLV was computed as: where is the total number of temporal samples. where is the number of temporal samples. This equation represents the MVL for two distinct 518 single frequencies ( , ) chosen within the gamma and theta bandwidths, respectively. To 519 illustrate the phase-amplitude coupling of the two oscillatory bands, we plotted 520 comodulograms, which are 2D plots that specify the MVL between different pairs of 521 frequencies with the low frequency ( ) along the x-axis and the high frequency ( ) along the 522 y-axis, and considering a frequency step size of 1Hz for each. To indicate the coupling strength between the two oscillatory bands with a single number, first, The datasets generated and analyzed in the current study are available at the OpenNeuro 551 repository. The data used for the main study is available at: 10.18112/openneuro.ds003800.v1.0.0 The data of the multi-sensory experiment is available at: 10.18112/openneuro.ds003805.v1.0.0 554 Code availability 555 The codes for generating the results presented in the manuscript are available at: 556 https://github.com/m-l-3/Gamma-Entrainment individuals safely induces highly coordinated 40 hz neural oscillation: A preliminary 640 study of non-invasive sensory stimulation for treating Alzheimer's disease. The road to restoring neural circuits for the 559 treatment of Alzheimer's disease 561 tangles, and Alzheimer's disease Alzheimer's Disease and the β-Amyloid Peptide Network abnormalities and interneuron dysfunction in 566 Alzheimer disease Inhibitory interneuron deficit links altered network activity and 568 cognitive dysfunction in alzheimer model Human EEG gamma oscillations in neuropsychiatric 570 disorders Gamma oscillations in cognitive disorders Olfactory response as a marker for Alzheimer's disease: 574 Evidence from perceptual and frontal lobe oscillation coherence deficit Remodeling of Inhibitory Hippocampal Circuits in Mouse Models of Alzheimer's 578 Disease Epilepsy and cognitive impairments in alzheimer disease Neuronal activity regulates the regional vulnerability to amyloid-β 582 deposition Cognitive tasks augment gamma EEG power The Theta-Gamma Neural Code Theta-Gamma coupling and working memory in Alzheimer's 589 dementia and mild cognitive impairment Alterations of Coherent Theta and Gamma Network Oscillations as 591 an Early Biomarker of Temporal Lobe Epilepsy and Alzheimer's Disease. Front Cortical odor processing in health and disease FDA's Decision to Approve New Treatment for Alzheimer's Disease | 596 FDA Alzheimer's disease beyond amyloid: Can the repetitive 599 failures of amyloid-targeted therapeutics inform future approaches to dementia drug 600 discovery? Failure to demonstrate efficacy of 602 aducanumab: An analysis of the EMERGE and ENGAGE trials as reported by Biogen Prescribing Aducanumab in the Face of Meager 605 Efficacy and Real Risks Clinical Efficacy, Drug Safety, and Surrogate Endpoints: Has Aducanumab Met All of Its Expectations? Gamma frequency entrainment attenuates amyloid load and 610 modifies microglia Optogenetic gamma stimulation rescues memory impairments in an 612 Alzheimer's disease mouse model Alterations in hippocampal network oscillations and theta-gamma 614 coupling arise before Aβ overproduction in a mouse model of Alzheimer's disease Modulating Human Memory via 617 Entrainment of Brain Oscillations Causal Evidence 619 for a Role of Theta and Alpha Oscillations in the Control of Working Memory Multi-sensory Gamma Stimulation Ameliorates Alzheimer's-622 Gamma Entrainment Binds Higher-Order Brain Regions and Offers 624 Gamma Entrainment: Impact on Neurocircuits, Glia, and 626 Therapeutic Opportunities Effects of Rhythmic Sensory Stimulation in Alzheimer's Disease: An Exploratory 629 Pilot Study Gamma Band Light Stimulation in Human Case Studies: Groundwork 631 for Potential Alzheimer's Disease Treatment Gamma Entrainment in a Large 633 Retrospective Cohort: Implications for Photic Stimulation Therapy for Alzheimer's 634 Alzheimer's Dis Entrainment 636 of gamma oscillations by auditory chirp stimulation in Alzheimer's disease patients Sensory gamma frequency stimulation in cognitively healthy and AD 35 Dementia Patients: Results of a Preliminary Clinical Trial Induction of specific brain oscillations may restore neural circuits and 646 be used for the treatment of Alzheimer's disease Gamma Band Neural Stimulation in Humans and the Promise of 649 a New Modality to Prevent and Treat Alzheimer's Disease Stimulation Induces a Neuroimmune Signaling Profile Distinct from Acute 653 Synchronizing Brain Rhythms to 655 Improve Cognition EEG Features of Working Memory Impairment in Alzheimer's Dementia and Mild Cognitive Impairment Cross-frequency coupling based neuromodulation 659 for treating neurological disorders Reconstruction of functional brain network in Alzheimer's disease via 661 cross-frequency phase synchronization Cross-frequency coupling supports multi-item working memory in 663 the human hippocampus Hippocampal-prefrontal theta-gamma coupling during performance of a spatial 666 working memory task Theta-gamma coupling in the entorhinal-hippocampal system Different theta frameworks coexist in the rat hippocampus 670 and are coordinated during memory-guided and novelty tasks Optogenetic activation of parvalbumin and somatostatin interneurons 672 selectively restores theta-nested gamma oscillations and oscillation-induced spike 673 timing-dependent long-Term potentiation impaired by amyloid β oligomers Theta-Nested Gamma Oscillations in Next 678 Intrinsic cornu ammonis area 1 theta-nested gamma oscillations induced by optogenetic theta 681 frequency stimulation Comparison of three gamma oscillations in the 683 mouse entorhinal-hippocampal system Memory entrainment by visually evoked theta-685 gamma coupling Entrainment of Theta Oscillations in Sensory Regions Predicts Human Associative 688 Memory Performance A Natural Stimulus for Entraining 690 Brain Oscillations in Alzheimer's Disease Therapy: Canary Song. in Alzheimer's 691 Association International Conference Coupling Explains the Advantage of Auditory Plus Visual Gamma Entrainment in 694 Alzheimer's Therapy Different regional patterns of cortical thinning in Alzheimer's disease 696 and frontotemporal dementia The diagnosis of dementia due to Alzheimer's disease: 698 Recommendations from the National Institute on Aging-Alzheimer's Association 699 workgroups on diagnostic guidelines for Alzheimer's disease. Alzheimer's Dement Makoto's preprocessing pipeline -SCCN BrainNet Viewer: A Network Visualization Tool for 705 Human Brain Connectomics Quantification of phase-amplitude 707 coupling in neuronal oscillations:comparison of phase-locking value, mean vector 708 length, modulation index, and generalized-linear-modeling-cross-frequency-coupling Time-Frequency Based Phase-Amplitude Coupling The authors declare no competing interests.