Connectivity Metrics Research Articles

Integrating EEG and Machine Learning to Analyze Brain Changes during the Rehabilitation of Broca's Aphasia.

The fusion of electroencephalography (EEG) with machine learning is transforming rehabilitation. Our study introduces a neural network model proficient in distinguishing pre- and post-rehabilitation states in patients with Broca's aphasia, based on brain connectivity metrics derived from EEG recordings during verbal and spatial working memory tasks. The Granger causality (GC), phase-locking value (PLV), weighted phase-lag index (wPLI), mutual information (MI), and complex Pearson correlation coefficient (CPCC) across the delta, theta, and low- and high-gamma bands were used (excluding GC, which spanned the entire frequency spectrum). Across eight participants, employing leave-one-out validation for each, we evaluated the intersubject prediction accuracy across all connectivity methods and frequency bands. GC, MI theta, and PLV low-gamma emerged as the top performers, achieving 89.4%, 85.8%, and 82.7% accuracy in classifying verbal working memory task data. Intriguingly, measures designed to eliminate volume conduction exhibited the poorest performance in predicting rehabilitation-induced brain changes. This observation, coupled with variations in model performance across frequency bands, implies that different connectivity measures capture distinct brain processes involved in rehabilitation. The results of this paper contribute to current knowledge by presenting a clear strategy of utilizing limited data to achieve valid and meaningful results of machine learning on post-stroke rehabilitation EEG data, and they show that the differences in classification accuracy likely reflect distinct brain processes underlying rehabilitation after stroke.

A novel recognition and classification approach for motor imagery based on spatio-temporal features.

Motor imagery, as a paradigm of brainmachine interfaces, holds vast potential in the field of medical rehabilitation. Addressing the challenges posed by the non-stationarity and low signal-to-noise ratio of EEG signals, the effective extraction of features from motor imagery signals for accurate recognition stands as a key focus in motor imagery brain-machine interface technology. This paper proposes a motor imagery EEG signal classification model that combines functional brain networks with graph convolutional networks. First, functional brain networks are constructed using different brain functional connectivity metrics, and graph theory features are calculated to deeply analyze the characteristics of brain networks under different motor tasks. Then, the constructed functional brain networks are combined with graph convolutional networks for the classification and recognition of motor imagery tasks. The analysis based on brain functional connectivity reveals that the functional connectivity strength during the both fists task is significantly higher than that of other motor imagery tasks, and the functional connectivity strength during actual movement is generally superior to that of motor imagery tasks. In experiments conducted on the Physionet public dataset, the proposed model achieved a classification accuracy of 88.39% under multi-subject conditions, significantly outperforming traditional methods. Under single-subject conditions, the model effectively addressed the issue of individual variability, achieving an average classification accuracy of 99.31%. These results indicate that the proposed model not only exhibits excellent performance in the classification of motor imagery tasks but also provides new insights into the functional connectivity characteristics of different motor tasks and their corresponding brain regions.

Autism Spectrum Disorder Detection by Hybrid Convolutional Recurrent Neural Networks from Structural and Resting State Functional MRI Images.

This study aims to increase the accuracy of autism spectrum disorder (ASD) diagnosis based on cognitive and behavioral phenotypes through multiple neuroimaging modalities. We apply machine learning (ML) algorithms to classify ASD patients and healthy control (HC) participants using structural magnetic resonance imaging (s-MRI) together with resting state functional MRI (rs-f-MRI and f-MRI) data from the large multisite data repository ABIDE (autism brain imaging data exchange) and identify important brain connectivity features. The 2D f-MRI images were converted into 3D s-MRI images, and datasets were preprocessed using the Montreal Neurological Institute (MNI) atlas. The data were then denoised to remove any confounding factors. We show, by using three fusion strategies such as early fusion, late fusion, and cross fusion, that, in this implementation, hybrid convolutional recurrent neural networks achieve better performance in comparison to either convolutional neural networks (CNNs) or recurrent neural networks (RNNs). The proposed model classifies subjects as autistic or not according to how functional and anatomical connectivity metrics provide an overall diagnosis based on the autism diagnostic observation schedule (ADOS) standard. Our hybrid network achieved an accuracy of 96% by fusing s-MRI and f-MRI together, which outperforms the methods used in previous studies.

3D-FGAN: A 3D stochastic reconstruction method of digital cores

Digital core reconstruction can quantitatively characterize the interior microstructure of porous media and revealing variations in the petrophysical properties of rocks, which can be classified into physical experimental methods and numerical reconstruction methods. Numerical reconstruction methods utilize the information contained in 2D or 3D images to reconstruct 3D digital cores, which is relatively inexpensive and more efficient, but traditional numerical methods usually cannot reuse extracted statistical information from previous reconstruction process and may generate unsatisfying results. Nowadays, the development of deep learning has opened a door to the high-quality reconstruction of digital cores. Generative adversarial networks (GANs), viewed as an important branch of deep learning, have been used for the 3D digital core reconstruction. GANs have a strong feature extraction capability and can generate fake images that closely resemble the real ones. Nevertheless, due to the complex structure of 3D core images, especially those with high anisotropy, GANs and their variants still have difficulties in generating satisfying pore structure. To address the above issues, we propose a variant of GANs called 3D fast generative adversarial network (3D-FGAN) to rapidly generate high-quality 3D digital core images. Several channel attention modules called skip-layer excitation blocks are added to the generator of 3D-FGAN, enhancing the utilization of important features. Auto-encoders are employed to optimize the discriminator of 3D-FGAN, which makes the discriminator more comprehensively extract the features of input images and thus better guides the generator. Through the evaluation metrics of pore space distribution (e.g., isolated pore space and connected pore space, pore diameters and throat diameters), multiple-point connectivity, absolute permeability and diversity test, the quality of 3D-FGAN can be proved. Meanwhile, in the reconstruction of anisotropic shales with evident fractures, 3D-FGAN can quickly produce higher-quality images compared with some other numerical reconstruction methods.

Investigating the influence of integrated landscape connectivity on water dissolved organic carbon variations in farm pond catchments of subtropical southern China

The landscape connectivity in farm pond catchments leads to complex hydrologic and biogeochemical processes, which is important to the carbon cycle of the watershed. However, little attention has been paid to analyzing the influence of landscape connectivity on water dissolved organic carbon (DOC) content in farm pond catchments. In this study, we developed an integrated landscape connectivity framework that encompasses landscape connectivity metrics, graph theory connectivity indices, and spatial distance. And the graph theory connectivity indices were constructed based on the modified resistance surface using a function that integrated landscape development intensity, slope, and livestock density. Using DOC content data from 86 farm ponds monitored between June 2017 and May 2018. The results demonstrated that the average water DOC content in farm ponds was higher during the dry season (5.35 ± 2.03 mg/L) compared to the wet season (4.27 ± 1.40 mg/L). The individual effects of landscape connectivity metrics and graph theory connectivity indices on DOC content were more prominent during the dry season, while the spatial distance had a more significant influence in the wet season. The interactive effects of determinant variables had a more substantial influence on pond water DOC content than their individual effects during both wet and dry seasons. The interactive effect variation ratio (IEVR) of all variables was positive in both seasons (0.143–0.777). Combining landscape connectivity metrics, graph theory connectivity indices, and spatial distance (22.0–24.3%) improved the interpretation of DOC content in comparison to using separate (0.4–12.1%) or paired variables (0.3–19.3%). Our findings suggest that integrated landscape connectivity is necessary to further understand the role of ponds in the organic carbon cycling of surface water. These strategies should not only focus on landscape connectivity metrics, but also consider the influences of graph theory connectivity indices and spatial distance in farm pond catchments.

Prediction of transportation index for urban patterns in small and medium-sized Indian cities using hybrid RidgeGAN model

The rapid urbanization trend in most developing countries including India is creating a plethora of civic concerns such as loss of green space, degradation of environmental health, scarcity of clean water, rise in air pollution, and exacerbated traffic congestion resulting in significant delays in vehicular transportation. To address the intricate nature of transportation issues, many researchers and planners have analyzed the complexities of urban and regional road systems using transportation models by employing transportation indices such as road length, network density, accessibility, and connectivity metrics. This study addresses the complexities of predicting road network density for small and medium-sized Indian cities that come under the Integrated Development of Small and Medium Towns (IDSMT) project at a national level. A hybrid framework based on Kernel Ridge Regression (KRR) and the CityGAN model is introduced to predict network density using spatial indicators of human settlements. The major goal of this study is to generate hyper-realistic urban patterns of small and medium-sized Indian cities using an unsupervised CityGAN model and to study the causal relationship between human settlement indices (HSIs) and transportation index (network density) using supervised KRR for the real cities. The synthetic urban universes mimic Indian urban patterns and evaluating their landscape structures through the settlement indices can aid in comprehending urban landscape, thereby enhancing sustainable urban planning. We analyzed 503 real cities to find the actual relationship between the urban settlements and their road density. The nonlinear KRR model may help urban planners in deriving the network density for GAN-generated futuristic urban patterns through the settlement indicators. The proposed hybrid process, termed as RidgeGAN model, can gauge the sustainability of urban sprawl tied to infrastructure and transportation systems in sprawling cities. Analysis results clearly demonstrate the utility of RidgeGAN in predicting network density for different kinds of human settlements, particularly for small and medium Indian cities. By predicting future urban patterns, this study can help in the creation of more livable and sustainable areas, particularly by improving transportation infrastructure in developing cities.

Percentage of area protected can substitute for more complicated structural metrics when monitoring protected area connectivity

Protected area (PA) targets have become a hallmark of global conservation policies such as the recent Kunming-Montreal Global Biodiversity Framework, which requires signatory countries to establish new PAs while also monitoring gains to both PA network coverage and connectivity at large spatial scales. Policy makers tasked with implementing and monitoring progress toward these targets face the difficult decision about which connectivity metric to use, which is not always straightforward given existing data and software limitations. We empirically compare 17 of the most widely used structural connectivity metrics to determine how they capture change in a PA network as additional protected areas are added and assess whether mathematically simple metrics are a reasonable substitute for more complex metrics. We find that simply reporting the percentage of the total area that is protected is a viable way to capture connectivity gains in most landscapes. If a more involved metric is desired, we recommend the Integral Index of Connectivity, which was highly correlated with the percentage of area protected, produced similar results when measuring change in PA networks, and incorporates stepping stone movements through transboundary regions.

Enhancing autism spectrum disorder classification in children through the integration of traditional statistics and classical machine learning techniques in EEG analysis

Autism Spectrum Disorder (ASD) is a neurodevelopmental disorder hallmarked by challenges in social communication, limited interests, and repetitive, stereotyped movements and behaviors. Numerous research efforts have indicated that individuals with ASD exhibit distinct brain connectivity patterns compared to control groups. However, these investigations, often constrained by small sample sizes, have led to inconsistent results, suggesting both heightened and diminished long-range connectivity within ASD populations. To bolster our analysis and enhance their reliability, we conducted a retrospective study using two different connectivity metrics and employed both traditional statistical methods and machine learning techniques. The concurrent use of statistical analysis and classical machine learning techniques advanced our understanding of model predictions derived from the spectral or connectivity attributes of a subject's EEG signal, while also verifying these predictions. Significantly, the utilization of machine learning methodologies empowered us to identify a unique subgroup of correctly classified children with ASD, defined by the analyzed EEG features. This improved approach is expected to contribute significantly to the existing body of knowledge on ASD and potentially guide personalized treatment strategies.

Proteins involved in synapse organization are associated with longitudinal loss in gray matter network connectivity in prodromal AD

AbstractBackgroundLower gray matter connectivity measures are related to faster cognitive decline in prodromal Alzheimer’s Disease (AD). We investigated the biological underpinnings responsible for this loss of brain connections using cerebrospinal fluid (CSF) proteomics.MethodData from ADNI was used to select 82 individuals with MCI and abnormal CSF amyloid (mean age±SD 74±7) and 37 controls with normal amyloid levels (mean age±SD 75±5), for whom repeated MRI‐scans (median of 6 scans over 1.5 follow‐up years) and baseline CSF proteomic data (in total 306 proteins) were available. We used an automated pipeline (Tijms, Series et al. 2012) to construct gray matter networks from 3D‐T1 sequences and calculate connectivity density and gamma – measures previously found to be most strongly related to cognitive decline (Dicks, Vermunt et al. 2020). Linear mixed models were applied to test associations between baseline CSF protein levels and repeated connectivity metrics, controlling for intracranial volume, age and sex. Proteins that demonstrated significant associations (p<0.05) were identified and examined for their biological relevance according to Gene Ontology.ResultConnectivity density declined over time with ‐0.8% per year (pMCI = 0.001; figure 1), which did not differ from controls (pdif = 0.15). Gamma declined faster in MCI than in controls (bMCI = ‐0.016 vs. bCN = ‐0.007, pdif <0.001; figure 1). Higher levels of 99/108 proteins were associated with faster decline in connectivity density (a.o. tau and neuroprentaxins). These proteins were involved in regulation of synapse organization (GO:0050807; pFDR = 7.00E‐07; figure 2). Higher levels of 9/108 proteins associated with less decline and were involved in chemotaxis (GO:0048247; pFDR = 1.40E‐02). Similarly, 25 out 30 proteins (83%) associated with alterations in gamma over time, showed associations of higher levels with less gamma decline and were also involved in chemotaxis (GO:0050926; pFDR = 9.42E‐05; figure 3). There were no protein‐connectivity relationships in controls.ConclusionProteins involved in synapse organization were associated with faster decline in connectivity density over time in patients, while chemotaxis was associated with less decline. This suggests that synaptic degeneration may predict disease progression while immune responses may serve as a resilience mechanism in AD.

Cerebellar‐cortical functional connectivity is associated with Alzheimer’s disease pathology and cerebellar atrophy

AbstractBackgroundBrain atrophy has been associated with Alzheimer’s disease (AD) pathology. Previous research also suggests that functional connectivity (FC) among distributed neural networks is disrupted in preclinical AD. The cerebellum has been thought to play a crucial role in motor control, yet recent evidence points to its additional role in non‐motor domains as well as cognitive decline in AD. Cerebellar volume has been shown to decline along the course of AD progression. Moreover, neuroimaging studies have delineated intrinsic cerebellar‐cortical functional networks that presumably support cognitive functions. Despite the accumulating findings on neurodegeneration or FC, how cerebellar FC and atrophy relate to AD pathology in the early stage has not been extensively investigated.MethodData from 83 cognitively normal older adults (mean 71.0 yrs, 66% females) in the Alzheimer’s Disease Neuroimaging Initiative were analyzed. Volumetric data were obtained from FreeSurfer 6.0 segmentation. Resting state fMRI scans were preprocessed using the CONN Toolbox. Aß and tau standardized uptake value ratio (SUVR) data were obtained from [18F]‐Florbetapir or [18F]‐Florbetaben (Aß) and [18F]‐Flortaucipir (tau) PET scans. For Aß, SUVR values were transformed to the Centiloid scale, and a global summary measure was used. For tau, SUVR values in the temporal meta regions of interest were used. Seed‐based functional connectivity metrics were computed using the 7 cerebellar parcellations (Buckner et al., 2011) that are functionally coupled to different cortical networks (Yeo et al., 2011).ResultCerebellar volume showed no significant relationship with amyloid or tau SUVR. By contrast, hippocampal atrophy was positively associated with amyloid or tau SUVR and showed trending significance (ps < 0.1). Increased Aß burden was associated with increased cerebellar‐cortical FC in the attention, frontoparietal, and default mode networks. Increased tau load was associated with increased cerebellar‐cortical FC in the visual, somatomotor, and attention networks. Cerebellar volume was associated with cerebellar‐cortical FC in the visual, somatomotor, attention networks, but not with the limbic, frontoparietal, or default mode networks.ConclusionOur findings suggest that functional cerebellar‐cortical connectivity varies with amyloid and tau pathology in different networks, and such functional changes may precede cerebellar atrophy in preclinical AD.

Cerebellar‐cortical functional connectivity is associated with Alzheimer’s disease pathology and cerebellar atrophy

AbstractBackgroundBrain atrophy has been associated with Alzheimer’s disease (AD) pathology. Previous research also suggests that functional connectivity (FC) among distributed neural networks is disrupted in preclinical AD. The cerebellum has been thought to play a crucial role in motor control, yet recent evidence points to its additional role in non‐motor domains as well as cognitive decline in AD. Cerebellar volume has been shown to decline along the course of AD progression. Moreover, neuroimaging studies have delineated intrinsic cerebellar‐cortical functional networks that presumably support cognitive functions. Despite the accumulating findings on neurodegeneration or FC, how cerebellar FC and atrophy relate to AD pathology in the early stage has not been extensively investigated.MethodData from 83 cognitively normal older adults (mean 71.0 yrs, 66% females) in the Alzheimer’s Disease Neuroimaging Initiative were analyzed. Volumetric data were obtained from FreeSurfer 6.0 segmentation. Resting state fMRI scans were preprocessed using the CONN Toolbox. Aβ and tau standardized uptake value ratio (SUVR) data were obtained from [18F]‐Florbetapir or [18F]‐Florbetaben (Aβ) and [18F]‐Flortaucipir (tau) PET scans. For Aβ, SUVR values were transformed to the Centiloid scale, and a global summary measure was used. For tau, SUVR values in the temporal meta regions of interest were used. Seed‐based functional connectivity metrics were computed using the 7 cerebellar parcellations (Buckner et al., 2011) that are functionally coupled to different cortical networks (Yeo et al., 2011).ResultCerebellar volume showed no significant relationship with amyloid or tau SUVR. By contrast, hippocampal atrophy was positively associated with amyloid or tau SUVR and showed trending significance (ps < 0.1). Increased Aβ burden was associated with increased cerebellar‐cortical FC in the attention, frontoparietal, and default mode networks. Increased tau load was associated with increased cerebellar‐cortical FC in the visual, somatomotor, and attention networks. Cerebellar volume was associated with cerebellar‐cortical FC in the visual, somatomotor, attention networks, but not with the limbic, frontoparietal, or default mode networks.ConclusionOur findings suggest that functional cerebellar‐cortical connectivity varies with amyloid and tau pathology in different networks, and such functional changes may precede cerebellar atrophy in preclinical AD.

Dynamical irreversibility, high‐order interactions, and whole‐brain modeling of gamma oscillations in neurodegeneration

AbstractBackgroundAbnormal gamma oscillations (γ) have been systematically reported in preclinical animal models of Alzheimer’s disease (AD) and human AD patients. However, little is known about the underlying mechanisms, non‐linear dynamics, and high‐order interactions of γ in dementia.MethodTo bridge this gap, we combined EEG and fMRI with three novel approaches to offer a comprehensive pathophysiological characterization of γ oscillations: We assessed the irreversibility of large‐scale brain dynamics to provide a precise signature capable of distinguishing AD at the global, local and network levels and at different oscillatory regimes (study 1). A high‐order functional connectivity (HOFC) multivariate information theory approach capturing interactions between 3 or more regions across spatiotemporal scales was compared with standard connectivity metrics for dementia characterization (study 2). A whole‐brain semi‐empirical model with a perturbational approach combined EEG, structural connectivity, and atrophy to explain the disrupted high‐order interactions observed in AD (study 3).ResultsStudy 1: AD was associated with a breakdown of temporal irreversibility at the global, local, and network levels at γ and other oscillations. The irreversibility of brain dynamics provided higher accuracy and more distinctive information than classical neurocognitive measures when differentiating AD from controls. Study 2: EEG‐HOFC revealed hypoconnectivity in γ between frontal, limbic, and sensory regions in AD, and large effect sizes, compared to standard pairwise metrics, proving a more accurate and parsimonious characterization of AD and frontotemporal dementia (FTD) patients. Study 3: whole‐brain modeling showed that a combined mechanism involving reduced structural integration (DTI) and hypo‐excitability (altered excitation/inhibition balance) triggered the altered high‐order interaction γ dynamics observed in AD and FTD. A perturbational approach identified physiopathological regions (SFG, MTG, precuneus) deviating the trajectories from disease to a healthy brain state.ConclusionThe present results offer new evidence regarding the pathophysiological link, mechanistic processes, and non‐linear whole‐brain dynamics between altered resting‐state γ rhythms and the characterization of AD and FTD, opening new avenues for dementia research at different levels.

Functional connectivity moderates the impact of synaptic loss on behaviour in frontotemporal lobar degeneration syndromes

AbstractBackgroundThere is growing evidence for early and extensive synaptic loss in frontotemporal lobar degeneration syndromes from in vivo and post‐mortem studies. We proposed that synaptic loss would affect behaviour, in relation to changes in functional connectivity.MethodWe recruited 29 participants with progressive supranuclear palsy, 16 participants with amyloid‐negative corticobasal syndrome, 8 participants with behavioural variant frontotemporal dementia and 24 similarly‐aged healthy volunteers. All participants underwent resting‐state functional 3T MRI and positron emission tomography with [11C]UCB‐J, that binds synaptic vesicle 2A. We compared the connectivity metrics of nodal weighted degree and voxelwise regional homogeneity with [11C]UCB‐J non‐displaceable binding potential (BPND), in associative and commonality analyses. We then used independent component analysis to extract [11C]UCB‐J BPND signal variance across participants, followed by derivation of participant‐specific scores of functional spatial covariance with the [11C]UCB‐J components. We applied model selection to assess whether connectivity scores would improve prediction of and moderate the effect of synaptic loss on clinical severity.ResultReduced [11C]UCB‐J BPND was associated with reduced connectivity, with synaptic density predicting both measures of connectivity over and above grey matter volume (Figure 1). The anatomical distribution of synaptic loss partially overlapped in PSP, bvFTD and CBS, but there were also disease‐specific effects (Figure 2). The effect of synaptic loss on functional connectivity was expressed local to the regions of severe synaptic loss, and in remote regions. Moreover, functional connectivity explained additional variance and moderated the relationship between synaptic density and clinical severity (Figure 3).ConclusionIn vivo synaptic loss is associated with reduced connectivity, and changes in behaviour, with implications for modelling disease pathogenesis and translational studies.

Novel machine learning approaches for improving the reproducibility and reliability of functional and effective connectivity from functional MRI

Objective. New measures of human brain connectivity are needed to address gaps in the existing measures and facilitate the study of brain function, cognitive capacity, and identify early markers of human disease. Traditional approaches to measure functional connectivity (FC) between pairs of brain regions in functional MRI, such as correlation and partial correlation, fail to capture nonlinear aspects in the regional associations. We propose a new machine learning based measure of FC ( ML.FC ) which efficiently captures linear and nonlinear aspects. Approach. To capture directed information flow between brain regions, effective connectivity (EC) metrics, including dynamic causal modeling and structural equation modeling have been used. However, these methods are impractical to compute across the many regions of the whole brain. Therefore, we propose two new EC measures. The first, a machine learning based measure of effective connectivity ( ML.EC ), measures nonlinear aspects across the entire brain. The second, Structurally Projected Granger Causality ( SP.GC ) adapts Granger Causal connectivity to efficiently characterize and regularize the whole brain EC connectome to respect underlying biological structural connectivity. The proposed measures are compared to traditional measures in terms of reproducibility and the ability to predict individual traits in order to demonstrate these measures’ internal validity. We use four repeat scans of the same individuals from the Human Connectome Project and measure the ability of the measures to predict individual subject physiologic and cognitive traits. Main results. The proposed new FC measure of ML.FC attains high reproducibility (mean intra-subject R 2 of 0.44), while the proposed EC measure of SP.GC attains the highest predictive power (mean R 2 across prediction tasks of 0.66). Significance. The proposed methods are highly suitable for achieving high reproducibility and predictiveness and demonstrate their strong potential for future neuroimaging studies.

EEG network functional connectivity during sleep and in presence of network hyperexcitability in Alzheimer’s disease

AbstractBackgroundFunctional connectivity (FC) studies in Alzheimer’s disease (AD) have focused primarily on the awake resting state. Yet, sleep disturbances in AD are common and occur early in the disease. Network hyperexcitability in AD also arises during sleep, resulting in epileptiform activity and seizures, and impacting memory consolidation. Hence, understanding the FC changes that occur during sleep in AD could provide important biological insights. To date, no studies have examined FC changes occurring during sleep in AD and their relationship with network hyperexcitability.MethodUsing ambulatory scalp EEG recordings, we assessed sleep FC in cognitively normal healthy controls (HC, n = 35), individuals with early clinical stages of AD with no history or risk factors for epilepsy (ADNoEp, n = 44), and individuals with early clinical stages of AD and epilepsy related to AD (ADEp, n = 14). We measured FC using corrected amplitude envelope correlation and assessed FC on different scales, including channel‐pair associations, regional associations, seed‐based connectivity, and graph theory metrics. Machine learning classifiers were developed to determine whether sleep FC measures provide accurate discrimination between HC, ADNoEp, and ADEp.ResultThere were no significant differences in age, sex, or education between HC, ADNoEp, and ADEp. Compared to HC, the ADNoEp group had 1) reduced FC during REM sleep in the alpha band in posterior regions; and 2) increased FC during NREM2 sleep in the gamma‐band in frontal regions. Compared to ADNoEp, the ADEp group had reduced FC in NREM2 sleep in the delta‐band in temporal and posterior regions. Graph theory metrics corroborated these sleep stage‐ and frequency‐specific findings between groups. Machine learning classifiers using graph theory metrics as input features distinguished ADNoEp from HC, and ADNoEp from ADEp, with high performance (AUC>0.85).ConclusionAD is associated with sleep stage‐, frequency band‐, and region‐specific changes in FC. Network hyperexcitability in ADEp is also associated with changes in sleep FC, which are distinct from those seen in ADNoEp. Our findings expand understanding of brain network changes in AD to the sleep state and demonstrate sleep‐stage specific alterations in FC that may have important implications for understanding AD neurophysiology and for disease monitoring and therapeutic development.

Early, low-dose hydrocortisone and near-term brain connectivity in extremely preterm infants.

Postnatal steroids are used to prevent bronchopulmonary dysplasia in extremely preterm infants but may have adverse effects on brain development. We assessed connectivity metrics of major cerebral and cerebellar white matter pathways at near-term gestational age among infants who did or did not receive a standardized regimen of hydrocortisone during the first 10 days of life. Retrospective cohort study. Infants born <28 weeks: Protocol group (n = 33) received at least 50% and not more than 150% of an intended standard dose of 0.5 mg/kg hydrocortisone twice daily for 7 days, then 0.5 mg/kg per day for 3 days; Non-Protocol group (n = 22), did not receive protocol hydrocortisone or completed <50% of the protocol dose. We assessed group differences in near-term diffusion MRI mean fractional anisotropy (FA) and mean diffusivity (MD) across the corticospinal tract, inferior longitudinal fasciculus, corpus callosum and superior cerebellar peduncle. Groups were comparable in gestational age, post-menstrual age at scan, medical complications, bronchopulmonary dysplasia, and necrotizing enterocolitis. No significant large effect group differences were identified in mean FA or MD in any cerebral or cerebellar tract. Low dose, early, postnatal hydrocortisone was not associated with significant differences in white matter tract microstructure at near-term gestational age. This study compared brain microstructural connectivity as a primary outcome among extremely preterm infants who did or did not receive early postnatal hydrocortisone. Low dose hydrocortisone in the first 10 days of life was not associated with significant differences in white matter microstructure in major cerebral and cerebellar pathways. Hydrocortisone did not have a significant effect on early brain white matter circuits.

A new strategy for optimizing marine protection networks by considering functional connectivity: An example of Tachypleus tridentatus

The concept of creating a conservation network through the establishment of multiple protected areas to protect marine fauna is gaining momentum. To effectively establish a network that yields favorable outcomes, it is imperative to identify suitable areas that foster optimal connectivity. Given the swift development of coastal areas, it has become necessary to develop cost-effective measures that yield maximum conservation benefits. However, exploration of these ideas for application to marine conservation networks remains scarce. In this study, we used the horseshoe crab (Tachypleus tridentatus) in the Beibu Gulf as an example. The species and environmental data were collected through field sampling and remote sensing. The species distribution model, the least-cost path model, and circuit theory were employed to establish a marine conservation network. The functional connectivity metrics were employed to optimize both the interconnection and economic efficiency of the network. The results indicated a low presence of adult horseshoe crabs in the Beibu Gulf, with average densities of 1.47 and 1.43 ind./km2 among the 38 surveyed stations in 2018 and 2019, respectively. The species distribution model accurately predicted juvenile presence in mudflats and adult presence offshore, and it identified 43 potential areas that are suitable for the establishment of marine protected areas. The marine protected areas that ranked higher in importance contributed significantly to maintaining the integrity of regional connectivity. The network connectivity shows a law of diminishing returns as the number of marine protected areas increases. The framework developed in this study will provide decision support for building and optimizing marine conservation networks.

Unilateral Mitochondrial-Hemodynamic Coupling and Bilateral Connectivity in the Prefrontal Cortices of Young and Older Healthy Adults.

A recent study demonstrated that noninvasive measurements of cortical hemodynamics and metabolism in the resting human prefrontal cortex can facilitate quantitative metrics of unilateral mitochondrial-hemodynamic coupling and bilateral connectivity in infraslow oscillation frequencies in young adults. The infraslow oscillation includes three distinct vasomotions with endogenic (E), neurogenic (N), and myogenic (M) frequency bands. The goal of this study was to prove the hypothesis that there are significant differences between young and older adults in the unilateral coupling (uCOP) and bilateral connectivity (bCON) in the prefrontal cortex. Accordingly, we performed measurements from 24 older adults (67.2 ± 5.9 years of age) using the same two-channel broadband near-infrared spectroscopy (bbNIRS) setup and resting-state experimental protocol as those in the recent study. After quantification of uCOP and bCON in three E/N/M frequencies and statistical analysis, we demonstrated that older adults had significantly weaker bilateral hemodynamic connectivity but significantly stronger bilateral metabolic connectivity than young adults in the M band. Furthermore, older adults exhibited significantly stronger unilateral coupling on both prefrontal sides in all E/N/M bands, particularly with a very large effect size in the M band (>1.9). These age-related results clearly support our hypothesis and were well interpreted following neurophysiological principles. The key finding of this paper is that the neurophysiological metrics of uCOP and bCON are highly associated with age and may have the potential to become meaningful features for human brain health and be translatable for future clinical applications, such as the early detection of Alzheimer's disease.

Efficient analysis of hydrological connectivity using 1D and 2D Convolutional Neural Networks

Understanding hydrological connectivity is essential to investigate ecological processes in river catchments and floodplains. Assessing flooding behavior, including flooded areas and connection times, is required to analyze hydrological connectivity in river floodplains. Deep learning, especially Convolutional Neural Networks (CNNs), is an attractive alternative to hydrodynamic modeling, which is more computationally expensive. This paper aims to develop a methodology to analyze the functional connectivity in remote and field measurement data-scarce areas using remote sensing data, CNN models, and connectivity metrics. The northern Lakes of the Narran River catchment, located in the Condamine–Balonne River floodplain in New South Wales, Australia, is the showcase for this method. One-dimensional CNN and two-dimensional U-Net configurations were applied and yielded comparable flood extents to the satellite images with Hit Rate values of 0.853 and 0.873, respectively. Two algorithms for determining hydrological connectivity were investigated, including the geostatistical Connectivity Function (CF) and the newly proposed Potential Connection Length (PCL). It was found that the connection along the main Narran River stream was more substantial than between the river and the floodplain lakes. The analysis using the PCL shows that the connectivity patterns in different stages of a flood event can vary depending on the initial condition of the floodplain. The overall conclusion from this work is that hydrological connectivity can be assessed computationally efficiently using only remote sensing, discharge data, and CNN models.

Coco: conservation design for optimal ecological connectivity

Despite global conservation efforts, biodiversity continues to decline, causing many species to face extinction. These efforts include designing protected areas to function as ecologically connected networks for habitat and movement pathway conservation. Ecological connectivity is defined as the connectivity of landscapes and seascapes that allows species to move and ecological processes to function unimpeded. It facilitates long-term species persistence and resilience, mitigates the impact of habitat fragmentation due to climate change and land-use change, and addresses ecological processes that support ecosystems. Thus, ecological connectivity is key in the design of habitat conservation networks. To incorporate many complicating factors in this process, it relies on decision-support frameworks to decide which areas to include to protect biodiversity while minimizing cost. Various approaches emerged to deal with the computational complexity involved in habitat conservation design. However, despite the importance of designing ecologically connected conservation networks, these widely used decision-support frameworks do not offer functionality to optimize ecological connectivity directly during conservation design. Here, we present a fast, exact method to use connectivity metrics during the biodiversity conservation design process. Our method is exact in the sense that it always returns optimal solutions in our model. We extend an existing Reserve Selection problem (RSP) formulation with vertex-weighted connectivity constraints to include edge-weighted connectivity constraints. Further, we describe two novel variations of the RSP to directly optimize over connectivity metrics, one with cost minimization and one with a fixed cost. We introduce Coco, an open-source decision-support system to design ecologically connected conservations. Coco provides an integer linear programming (ILP) method to include connectivity in conservation design. To this end, we formulate our novel RSP variations as an ILP. We test Coco on simulated data and two real datasets, one dataset of the Great Barrier Reef and a large-scale dataset of the marine area in British Columbia. We compare the performance of Coco to Marxan Connect and show that Coco outperforms Marxan Connect both in runtime and solution quality. Further, we compare the results of our proposed methods to the existing RSP formulation and show that our novel methods significantly increase connectivity at a lower cost.