Direct shear behaviors of excavated clay reinforced with geocomposite drainage layer encapsulated in thin sand layers.

The objective of this study is to assess the effectiveness of a novel structure comprising a geocomposite drainage layer and a thin sand layer (GDL + sand) in mitigating the rapid dumping of excavated clay and its associated issues, such as landslides. Two sets of direct shear tests were conducted to investigate the influence of sand layer thickness and compaction degree on the interface shear behavior of the GDL + sand structure. As the sand layer thickness increased, both the interface shear strength and friction angle gradually increased, first more sharply and then at a slower rate toward stability, while the interface cohesion decreased gradually. The optimal sand layer thickness for achieving the most effective reinforcement in stabilizing the clay was identified as 10 mm. A higher sand layer compaction degree was found to result in increased interface shear strength, interface friction angle, and interface cohesion. Building on these findings, the reinforcing efficiency of the GDL + sand structure was investigated through mechanism analysis in comparison to that of a geogrid + sand structure and GDL structure as per the interface friction coefficient. The ranking of interface friction coefficients among the three structures emerged as: geogrid + sand > GDL + sand > GDL. These results suggests that the GDL + sand structure exhibits superior reinforcement efficiency compared to the GDL structure and offers better drainage efficiency than the geogrid + sand structure.

Towards low-temperature dendrite-free zinc anode by constructing functional MXene buffer layer with duplex zincophilic sites.

Dendrite growth and side reactions of zinc metal anode have severely limited the practical application of aqueous zinc ion batteries (AZIBs). Herein, we introduce an artificial buffer layer composed of functional MXene (Ti3CN) for zinc anodes. The synthesized Ti3CN exhibits superior conductivity and features duplex zincophilic sites (N and F). These characteristics facilitate the homogeneous deposition of Zn2+, accelerate the desolvation process of hydrated Zn2+, and reduce the nucleation overpotential. The Ti3CN-protected Zn anode demonstrates significantly enhanced reversibility compared to bare Zn anode during long-term cycling, achieving a cumulative plating capacity of 10,000 mAh cm-2 at 10 mA cm-2. In Ti3CN-Zn||Cu asymmetric cell, it maintains nearly 100 % Coulombic efficiency over 2500 cycles at 2 mA cm-2. Furthermore, the assembled Ti3CN-Zn//δ-K0.51V2O5 (KVO) full cell exhibit a low capacity decay rate of 0.002 % per cycle at 5 A/g. Even at 0 °C, the Ti3CN-Zn symmetric cell maintains steady cycling for 2000 h. This study introduces a novel approach for designing artificial solid electrolyte interlayers for commercial AZIBs.

Temporal Alterations in White Matter in An App Knock-In Mouse Model of Alzheimer's Disease.

Alzheimer's disease (AD) is the most common form of dementia and results in neurodegeneration and cognitive impairment. White matter (WM) is affected in AD and has implications for neural circuitry and cognitive function. The trajectory of these changes across age, however, is still not well understood, especially at earlier stages in life. To address this, we used the AppNL-G-F/NL-G-F knock-in (APPKI) mouse model that harbors a single copy knock-in of the human amyloid precursor protein (APP) gene with three familial AD mutations. We performed in vivo diffusion tensor imaging (DTI) to study how the structural properties of the brain change across age in the context of AD. In late age APPKI mice, we observed reduced fractional anisotropy (FA), a proxy of WM integrity, in multiple brain regions, including the hippocampus, anterior commissure (AC), neocortex, and hypothalamus. At the cellular level, we observed greater numbers of oligodendrocytes in middle age (prior to observations in DTI) in both the AC, a major interhemispheric WM tract, and the hippocampus, which is involved in memory and heavily affected in AD, prior to observations in DTI. Proteomics analysis of the hippocampus also revealed altered expression of oligodendrocyte-related proteins with age and in APPKI mice. Together, these results help to improve our understanding of the development of AD pathology with age, and imply that middle age may be an important temporal window for potential therapeutic intervention.

Semi-quantitative study on the secondary compression characteristics of municipal solid waste in aerobic and anaerobic bioreactors.

Aeration plays a crucial role in accelerating the secondary compression of municipal solid waste (MSW) for the scientific implementation of aerobic bioreactor technology. There are few comparative reports on the secondary compaction characteristics of MSW in aerobic and anaerobic bioreactors. In this study, six long-term compression tests were conducted to analyze the impact of aeration on MSW compression characteristics, considering two degradation conditions (i.e. aerobic and anaerobic conditions) and three overburden stresses (i.e. 30, 50 and 100 kPa). Model-fitting analysis was employed to examine the data from the tests and exiting literatures. The results showed that aeration effectively increased the rate of secondary compression, and slightly enhanced the steady-state secondary compression strain. In addition, these enhancements tended to decrease with increasing stresses. The increment ratio of the secondary compression rate constant (Rk) was concentrated in the range of 25 % to 100 %, and increases with the increase of aeration rate. The increment ratio of the steady-state secondary compression strain (Rε) ranged from 10 % to 90 %, for the MSW with higher content of paper and wood exhibited higher Rε. The advance ratio of the secondary compression stabilization time (Rt) fell within the range of 20-50 %, and Rt is higher when the moisture content is in the range of 50-65 %. These findings provide valuable guidance on the accelerated stabilization in aerobic bioreactors, providing practical references for the application of aerobic technology to informal landfills.

Neighborhood Perceptions Are Associated With Intrinsic Amygdala Activity and Resting-State Connectivity With Salience Network Nodes Among Older Adults.

OBJECTIVE: Neighborhood perceptions are associated with physical and mental health outcomes; however, the biological associates of this relationship remain to be fully understood. Here, we evaluate the relationship between neighborhood perceptions and amygdala activity and connectivity with salience network (i.e., insula, anterior cingulate, thalamus) nodes. METHODS: Forty-eight older adults (mean age = 68 [7] years, 52% female, 47% non-Hispanic Black, 2% Hispanic) without dementia or depression completed the Perceptions of Neighborhood Environment Scale. Lower scores indicated less favorable perceptions of aesthetic quality, walking environment, availability of healthy food, safety, violence (i.e., more perceived violence), social cohesion, and participation in activities with neighbors. Participants separately underwent resting-state functional magnetic resonance imaging. RESULTS: Less favorable perceived safety ( ß = -0.33, pFDR = .04) and participation in activities with neighbors ( ß = -0.35, pFDR = .02) were associated with higher left amygdala activity, independent of covariates including psychosocial factors. Less favorable safety perceptions were also associated with enhanced left amygdala functional connectivity with the bilateral insular cortices and the left anterior insula ( ß = -0.34, pFDR = .04). Less favorable perceived social cohesion was associated with enhanced left amygdala functional connectivity with the right thalamus ( ß = -0.42, pFDR = .04), and less favorable perceptions about healthy food availability were associated with enhanced left amygdala functional connectivity with the bilateral anterior insula (right: ß = -0.39, pFDR = .04; left: ß = -0.42, pFDR = .02) and anterior cingulate gyrus ( ß = -0.37, pFDR = .04). CONCLUSIONS: Taken together, our findings document relationships between select neighborhood perceptions and amygdala activity as well as connectivity with salience network nodes; if confirmed, targeted community-level interventions and existing community strengths may promote brain-behavior relationships.

Monitoring the remediation of groundwater polluted by MSW landfill leachates by activated carbon and zeolite with spectral induced polarization technique.

The municipal solid waste (MSW) landfill in Hangzhou, China utilized zeolite and activated carbon (AC) as permeable reactive barrier (PRB) fill materials to remediate groundwater contaminated with MSW leachates containing ammonium, chemical oxygen demand (COD), and heavy metals. The spectral induced polarization (SIP) technique was chosen for monitoring the PRB because of its sensitivity to pore fluid chemistry and mineral-fluid interface composition. During the experiment, authentic groundwater collected from the landfill site was used to permeate two columns filled with zeolite and AC, and the SIP responses were measured at the inlet and outlet over a frequency range of 0.01-1000 Hz. The results showed that zeolite had a higher adsorption capacity for COD (7.08 mg/g) and ammonium (9.15 mg/g) compared to AC (COD: 2.75 mg/g, ammonium: 1.68 mg/g). Cation exchange was found to be the mechanism of ammonium adsorption for both zeolite and AC, while FTIR results indicated that π-complexation, π-π interaction, and electrostatic attraction were the main mechanisms of COD adsorption. The Cole-Cole model was used to fit the SIP responses and determine the relaxation time (τ) and normalized chargeability (mn). The calculated characteristic diameters of zeolite and AC based on the Schwarz equation and relaxation time (τ) matched the pore sizes observed from SEM and MIP, providing valuable information on contaminant distribution. The mn of zeolite was positively linear with adsorbed ammonium (R2 = 0.9074) and COD (R2 = 0.8877), while the mn of AC was negatively linear with adsorbed ammonium (R2 = 0.8192) and COD (R2 = 0.7916), suggesting that mn could serve as a surrogate for contaminant saturation. The laboratory-based real-time non-invasive SIP results showed good performance in monitoring saturation and provide a strong foundation for future field PRB monitoring.

Spectrum-Enhanced TRCA (SE-TRCA): A novel approach for direction detection in SSVEP-based BCI.

The Steady State Visual Evoked Potential (SSVEP) is a widely used component in BCIs due to its high noise resistance and low equipment requirements. Recently, a novel SSVEP-based paradigm has been introduced for direction detection, in which, unlike the common SSVEP paradigms that use several frequency stimuli, only one flickering stimulus is used and it makes direction detection very challenging. So far, only the CCA method has been used for direction detection using SSVEP component analysis. Since Canonical Correlation Analysis (CCA) has some limitations, a Task-Related Component Analysis (TRCA) based method has been introduced for feature extraction to improve the direction detection performance. Although these methods have been proven efficient, they do not utilize the latent frequency information in the EEG signal. Therefore, the performance of direction detection using SSVEP component analysis is still suboptimal. For further improvement, the TRCA-based algorithm is enhanced by incorporating frequency information and introducing Spectrum-Enhanced TRCA (SE-TRCA). SE-TRCA method can utilize frequency information in conjunction with spatial information by concatenating the EEG signal and its shifted version. Accordingly, the obtained spatio-spectral filters perform as a Finite Impulse Response (FIR) filter. To evaluate the proposed SE-TRCA method, two different sorts of datasets (1) a hybrid BCI dataset (including SSVEP component for direction detection) and (2) a pure benchmark SSVEP dataset (including SSVEP component for frequency detection) have been used. Our experiments showed that the accuracy of direction detection using the proposed SE-TRCA and TRCA approaches compared to CCA-based approach have been increased by 23.35% and 28.24%, respectively. Furthermore, the accuracy of character recognition obtained from integrating P300 and SSVEP components in CCA, TRCA, and SETRCA approaches are 54.01%, 56.02%, and 58.56%, on the hybrid dataset, respectively. The evaluation of the SE-TRCA method on the benchmark SSVEP dataset demonstrates that the SE-TRCA method outperforms both CCA and TRCA, particularly regarding frequency detection accuracy. In this specific dataset, the SE-TRCA method achieved an impressive frequency detection accuracy of 98.19% for a 3-s signal, surpassing the accuracies of TRCA and CCA, which were 97.91% and 90.47%, respectively. These results demonstrated that the TRCA-based approach is more efficient than the CCA approach to extracting spatial filters. Moreover, SE-TRCA, extracting both Spectral and spatial information from the EEG signal, can capture more discriminative features from the SSVEP component and increase the accuracy of classification. The results of this study emphasize the effectiveness of the proposed SE-TRCA approach across different SSVEP paradigms and tasks. These findings provide strong evidence for the method's ability to generalize well in SSVEP analysis.

One-dimensional consolidation of highly saturated abandoned-soil under time-dependent stress: An analytical study.

A thorough knowledge of the consolidation behavior of highly saturated soil under time-dependent stress is essential for the design and construction of abandoned-soil dump sites in the soft soil regions of China. In this study, one-dimensional consolidation analytical solutions are derived for such soil under one-way and two-way drainage conditions, accommodating the time-dependent stress created by various dumping protocols. Representative soil samples are obtained, and consolidation tests are conducted with various saturation degrees (one-way drainage) and loading protocols (two-way drainage), to verify the consolidation equation and determine its range of applicability to various saturation degrees. The effects of layer thickness, dumping type, and compaction degree on the consolidation behaviors of highly saturated abandoned-soil dumps are investigated. The one-dimensional consolidation equation is applicable to soil with saturation degree not lower than 75% under instantaneous stress, stepped stress, and linear stress. The pore pressure distribution with depth is not symmetrical; the eccentric distance of consolidation degree increases with increasing layer thickness in the stress application stage and is approximately zero in the stress keeping stage. The pore pressure at middle of the soil layer increases with increasing layer thickness and decreases with increasing dumping rate from the completion of soil dumping. With increasing compaction degree, the middle pore pressure increases, while the surface settlement decreases. In the premise of the stability of an abandoned-soil dump, where the goals are to reduce post-construction settlement and to shorten the consolidation process of the entire soil layer, the important factors are smaller layer thickness, higher dumping rate, and larger compaction degree.

Disrupted Excitation-Inhibition Balance in Cognitively Normal Individuals at Risk of Alzheimer's Disease.

BACKGROUND: Sex differences impact Alzheimer's disease (AD) neuropathology, but cell-to-network level dysfunctions in the prodromal phase are unclear. Alterations in hippocampal excitation-inhibition balance (EIB) have recently been linked to early AD pathology. OBJECTIVE: Examine how AD risk factors (age, APOEɛ4, amyloid-ß) relate to hippocampal EIB in cognitively normal males and females using connectome-level measures. METHODS: Individuals from the OASIS-3 cohort (age 42-95) were studied (N = 437), with a subset aged 65+ undergoing neuropsychological testing (N = 231). RESULTS: In absence of AD risk factors (APOEɛ4/Aß+), whole-brain EIB decreases with age more significantly in males than females (p = 0.021, ß= -0.007). Regression modeling including APOEɛ4 allele carriers (Aß-) yielded a significant positive AGE-by-APOE interaction in the right hippocampus for females only (p = 0.013, ß= 0.014), persisting with inclusion of Aß+ individuals (p = 0.012, ß= 0.014). Partial correlation analyses of neuropsychological testing showed significant associations with EIB in females: positive correlations between right hippocampal EIB with categorical fluency and whole-brain EIB with the Trail Making Test (p < 0.05). CONCLUSIONS: Sex differences in EIB emerge during normal aging and progresses differently with AD risk. Results suggest APOEɛ4 disrupts hippocampal balance more than amyloid in females. Increased excitation correlates positively with neuropsychological performance in the female group, suggesting a duality in terms of potential beneficial effects prior to cognitive impairment. This underscores the translational relevance of APOEɛ4 related hyperexcitation in females, potentially informing therapeutic targets or early interventions to mitigate AD progression in this vulnerable population.

Disrupted excitation-inhibition balance in cognitively normal individuals at risk of Alzheimer's disease.

Background: Sex differences impact Alzheimer's disease (AD) neuropathology, but cell-to-network level dysfunctions in the prodromal phase are unclear. Alterations in hippocampal excitation-inhibition balance (EIB) have recently been linked to early AD pathology. Objective: Examine how AD risk factors (age, APOE-ɛ4, amyloid-ß) relate to hippocampal EIB in cognitively normal males and females using connectome-level measures. Methods: Individuals from the OASIS-3 cohort (age 42-95) were studied (N = 437), with a subset aged 65+ undergoing neuropsychological testing (N = 231). Results: In absence of AD risk factors (APOE-ɛ4/Aß+), whole-brain EIB decreases with age more significantly in males than females (p = 0.021, ß = -0.007). Regression modeling including APOE-ɛ4 allele carriers (Aß-) yielded a significant positive AGE-by-APOE interaction in the right hippocampus for females only (p = 0.013, ß = 0.014), persisting with inclusion of Aß+ individuals (p = 0.012, ß = 0.014). Partial correlation analyses of neuropsychological testing showed significant associations with EIB in females: positive correlations between right hippocampal EIB with categorical fluency and whole-brain EIB with the trail-making test (p < 0.05). Conclusion: Sex differences in EIB emerge during normal aging and progresses differently with AD risk. Results suggest APOE-ɛ4 disrupts hippocampal balance more than amyloid in females. Increased excitation correlates positively with neuropsychological performance in the female group, suggesting a duality in terms of potential beneficial effects prior to cognitive impairment. This underscores the translational relevance of APOE-ɛ4 related hyperexcitation in females, potentially informing therapeutic targets or early interventions to mitigate AD progression in this vulnerable population.

Recent Progress for Concurrent Realization of Shuttle-Inhibition and Dendrite-Free Lithium-Sulfur Batteries.

Lithium-sulfur (Li-S) batteries have become one of the most promising new-generation energy storage systems owing to their ultrahigh energy density (2600 Wh kg-1 ), cost-effectiveness, and environmental friendliness. Nevertheless, their practical applications are seriously impeded by the shuttle effect of soluble lithium polysulfides (LiPSs), and the uncontrolled dendrite growth of metallic Li, which result in rapid capacity fading and battery safety problems. A systematic and comprehensive review of the cooperative combination effect and tackling the fundamental problems in terms of cathode and anode synchronously is still lacking. Herein, for the first time, the strategies for inhibiting shuttle behavior and dendrite-free Li-S batteries simultaneously are summarized and classified into three parts, including "two-in-one" S-cathode and Li-anode host materials toward Li-S full cell, "two birds with one stone" modified functional separators, and tailoring electrolyte for stabilizing sulfur and lithium electrodes. This review also emphasizes the fundamental Li-S chemistry mechanism and catalyst principles for improving electrochemical performance; advanced characterization technologies to monitor real-time LiPS evolution are also discussed in detail. The problems, perspectives, and challenges with respect to inhibiting the shuttle effect and dendrite growth issues as well as the practical application of Li-S batteries are also proposed.

Enhanced simulations of whole-brain dynamics using hybrid resting-state structural connectomes.

The human brain, composed of billions of neurons and synaptic connections, is an intricate network coordinating a sophisticated balance of excitatory and inhibitory activity between brain regions. The dynamical balance between excitation and inhibition is vital for adjusting neural input/output relationships in cortical networks and regulating the dynamic range of their responses to stimuli. To infer this balance using connectomics, we recently introduced a computational framework based on the Ising model, first developed to explain phase transitions in ferromagnets, and proposed a novel hybrid resting-state structural connectome (rsSC). Here, we show that a generative model based on the Kuramoto phase oscillator can be used to simulate static and dynamic functional connectomes (FC) with rsSC as the coupling weight coefficients, such that the simulated FC well aligns with the observed FC when compared to that simulated with traditional structural connectome. Simulations were performed using the open source framework The Virtual Brain on High Performance Computing infrastructure.

A convolutional neural network with pixel-wise sparse graph reasoning for COVID-19 lesion segmentation in CT images.

The COVID-19 pandemic has extremely threatened human health, and automated algorithms are needed to segment infected regions in the lung using computed tomography (CT). Although several deep convolutional neural networks (DCNNs) have proposed for this purpose, their performance on this task is suppressed due to the limited local receptive field and deficient global reasoning ability. To address these issues, we propose a segmentation network with a novel pixel-wise sparse graph reasoning (PSGR) module for the segmentation of COVID-19 infected regions in CT images. The PSGR module, which is inserted between the encoder and decoder of the network, can improve the modeling of global contextual information. In the PSGR module, a graph is first constructed by projecting each pixel on a node based on the features produced by the encoder. Then, we convert the graph into a sparsely-connected one by keeping K strongest connections to each uncertainly segmented pixel. Finally, the global reasoning is performed on the sparsely-connected graph. Our segmentation network was evaluated on three publicly available datasets and compared with a variety of widely-used segmentation models. Our results demonstrate that (1) the proposed PSGR module can capture the long-range dependencies effectively and (2) the segmentation model equipped with this PSGR module can accurately segment COVID-19 infected regions in CT images and outperform all other competing models.

Ordered Mesoporous Carbon Grafted MXene Catalytic Heterostructure as Li-Ion Kinetic Pump toward High-Efficient Sulfur/Sulfide Conversions for Li-S Battery.

Lithium-sulfur (Li-S) batteries exhibit unparalleled theoretical capacity and energy density than conventional lithium ion batteries, but they are hindered by the dissatisfactory "shuttle effect" and the sluggish conversion kinetics owing to the low lithium ion transport kinetics, resulting in rapid capacity fading. Herein, a catalytic two-dimensional heterostructure composite is prepared by evenly grafting mesoporous carbon on the MXene nanosheet (denoted as OMC-g-MXene), serving as interfacial kinetic accelerators in Li-S batteries. In this design, the grafted mesoporous carbon in the heterostructure can not only prevent the stack of MXene nanosheets with the enhanced mechanical property but also offer a facilitated pump for accelerating ion diffusion. Meanwhile, the exposed defect-rich OMC-g-MXene heterostructure inhibits the polysulfide shuttling with chemical interactions between OMC-g-MXene and polysulfides and thus simultaneously enhances the electrochemical conversion kinetics and efficiency, as fully investigated by in situ/ex situ characterizations. Consequently, the cells with OMC-g-MXene ion pumps achieve a high cycling capacity (966 mAh g-1 at 0.2 C after 200 cycles), a superior rate performance (537 mAh g-1 at 5 C), and an ultralow decaying rate of 0.047% per cycle after 800 cycles at 1 C. Even employed with a high sulfur loading of 7.08 mg cm-2 under lean electrolyte, an ultrahigh areal capacity of 4.5 mAh cm-2 is acquired, demonstrating a future practical application.

Detection of Mild Cognitive Impairment from Language Markers with Crossmodal Augmentation.

Mild cognitive impairment is the prodromal stage of Alzheimer's disease. Its detection has been a critical task for establishing cohort studies and developing therapeutic interventions for Alzheimer's. Various types of markers have been developed for detection. For example, imaging markers from neuroimaging have shown great sensitivity, while its cost is still prohibitive for large-scale screening of early dementia. Recent advances from digital biomarkers, such as language markers, have provided an accessible and affordable alternative. While imaging markers give anatomical descriptions of the brain, language markers capture the behavior characteristics of early dementia subjects. Such differences suggest the benefits of auxiliary information from the imaging modality to improve the predictive power of unimodal predictive models based on language markers alone. However, one significant barrier to the joint analysis is that in typical cohorts, there are only very limited subjects that have both imaging and language modalities. To tackle this challenge, in this paper, we develop a novel crossmodal augmentation tool, which leverages auxiliary imaging information to improve the feature space of language markers so that a subject with only language markers can benefit from imaging information through the augmentation. Our experimental results show that the multi-modal predictive model trained with language markers and auxiliary imaging information significantly outperforms unimodal predictive models.

Signed graph representation learning for functional-to-structural brain network mapping.

MRI-derived brain networks have been widely used to understand functional and structural interactions among brain regions, and factors that affect them, such as brain development and diseases. Graph mining on brain networks can facilitate the discovery of novel biomarkers for clinical phenotypes and neurodegenerative diseases. Since brain functional and structural networks describe the brain topology from different perspectives, exploring a representation that combines these cross-modality brain networks has significant clinical implications. Most current studies aim to extract a fused representation by projecting the structural network to the functional counterpart. Since the functional network is dynamic and the structural network is static, mapping a static object to a dynamic object may not be optimal. However, mapping in the opposite direction (i.e., from functional to structural networks) are suffered from the challenges introduced by negative links within signed graphs. Here, we propose a novel graph learning framework, named as Deep Signed Brain Graph Mining or DSBGM, with a signed graph encoder that, from an opposite perspective, learns the cross-modality representations by projecting the functional network to the structural counterpart. We validate our framework on clinical phenotype and neurodegenerative disease prediction tasks using two independent, publicly available datasets (HCP and OASIS). Our experimental results clearly demonstrate the advantages of our model compared to several state-of-the-art methods.

A first-order kinetic model for simulating the aerobic degradation of municipal solid waste.

Aerobic degradation models are important tools for investigating the aerobic degradation behavior of municipal solid waste (MSW). In this paper, a first-order kinetic model for aerobic degradation of MSW was developed. The model comprehensively considers the aerobic degradation of five substrates, i.e., holocellulose, non-cellulosic sugars, proteins, lipids and lignin. The proportion ranges of the five substrates are summarized with the recommended values given. The effects of temperature, moisture content, oxygen concentration and free air space (FAS) on the reaction rates are considered, and the effect of settlement is accounted for in the FAS correction function. The reliability of the model was verified by comparing simulations of the aerobic degradation of low food waste content (LFWC-) and high food waste content (HFWC-) MSWs to the literature. Afterwards, a sensitivity analysis was carried out to establish the relative importance of aeration rate (AR), volumetric moisture content (VMC), and temperature. VMC had the greatest influence on the aerobic degradation of LFWC-MSW, followed by temperature and then AR; for HFWC-MSW, temperature was the most important factor, then VMC and last was AR. The degradation ratio of LFWC-MSW can reach 98.0% after 100 days degradation under its optimal conditions (i.e., temperature: 55 °C, VMC: 40%, AR: 0.16 L min-1 kg-1 DM), while it is slightly higher as 99.5% for HFWC-MSW under its optimal conditions (i.e., temperature: 55 °C, VMC: 40%, AR: 0.20 L min-1 kg-1 DM).

BrainGB: A Benchmark for Brain Network Analysis With Graph Neural Networks.

Mapping the connectome of the human brain using structural or functional connectivity has become one of the most pervasive paradigms for neuroimaging analysis. Recently, Graph Neural Networks (GNNs) motivated from geometric deep learning have attracted broad interest due to their established power for modeling complex networked data. Despite their superior performance in many fields, there has not yet been a systematic study of how to design effective GNNs for brain network analysis. To bridge this gap, we present BrainGB, a benchmark for brain network analysis with GNNs. BrainGB standardizes the process by (1) summarizing brain network construction pipelines for both functional and structural neuroimaging modalities and (2) modularizing the implementation of GNN designs. We conduct extensive experiments on datasets across cohorts and modalities and recommend a set of general recipes for effective GNN designs on brain networks. To support open and reproducible research on GNN-based brain network analysis, we host the BrainGB website at https://braingb.us with models, tutorials, examples, as well as an out-of-box Python package. We hope that this work will provide useful empirical evidence and offer insights for future research in this novel and promising direction.

Bidirectional Mapping with Contrastive Learning on Multimodal Neuroimaging Data.

The modeling of the interaction between brain structure and function using deep learning techniques has yielded remarkable success in identifying potential biomarkers for different clinical phenotypes and brain diseases. However, most existing studies focus on one-way mapping, either projecting brain function to brain structure or inversely. This type of unidirectional mapping approach is limited by the fact that it treats the mapping as a one-way task and neglects the intrinsic unity between these two modalities. Moreover, when dealing with the same biological brain, mapping from structure to function and from function to structure yields dissimilar outcomes, highlighting the likelihood of bias in one-way mapping. To address this issue, we propose a novel bidirectional mapping model, named Bidirectional Mapping with Contrastive Learning (BMCL), to reduce the bias between these two unidirectional mappings via ROI-level contrastive learning. We evaluate our framework on clinical phenotype and neurodegenerative disease predictions using two publicly available datasets (HCP and OASIS). Our results demonstrate the superiority of BMCL compared to several state-of-the-art methods.