Browse Articles
Discover research articles across all indexed journals
Methionine metabolite spermidine inhibits tumor pyroptosis by enhancing MYO6-mediated endocytosis
Disrupted topological organization of brain connectome in patients with chronic low back related leg pain and clinical correlations
Abstract Chronic pain is associated with persistent alterations in brain structure and function. However, existing research has not fully explored the relationship between brain network topological properties and clinical symptoms in patients with chronic low back-related leg pain (cLBLP). In this study, we collected resting-state functional and structural magnetic resonance imaging data, along with clinical symptom evaluation data, from 32 cLBLP patients and 31 healthy controls. A large-scale complex network analysis was conducted to evaluate the global and nodal topological properties of functional and structural brain networks. Statistical analyses were performed to determine the associations between network properties and clinical variables. The results showed significant alterations in both global and nodal topological properties of functional and structural brain networks in cLBLP patients compared to healthy controls. Additionally, a direct correlation was found between structural network properties and spatial discrimination ability, measured by two-point tactile discrimination values, while no significant association was observed between functional connectivity and spatial discrimination. This study demonstrates that cLBLP patients exbibit a decreased local efficiency of functional connectivity network and increased compensatory global efficiency of structural connectivity network. Notably, alterations in the structural connectome, rather than the functional connectome, play a more significant role in deterioration of foot tactile spatial acuity in cLBLP patients. Trial registration : This trial was registered in the Chinese Clinical Trial Registry with the registration number ChiCTR2200055321 on 2022-01-06.
A framework for integrating genomics, microbial traits, and ecosystem biogeochemistry
Breast cancer prediction based on gene expression data using interpretable machine learning techniques
In-situ positive electrode-electrolyte interphase construction enables stable Ah-level Zn-MnO2 batteries
Generalized Gumbel model for r-largest order statistics, with an application to peak streamflow
Crucial roles of Grr1 in splicing and translation of HAC1 mRNA upon unfolded stress response
Fatigue analysis of an energy storage supercapacitor box under random vibration loading
Temporal Contrastive Learning through implicit non-equilibrium memory
Abstract The backpropagation method has enabled transformative uses of neural networks. Alternatively, for energy-based models, local learning methods involving only nearby neurons offer benefits in terms of decentralized training, and allow for the possibility of learning in computationally-constrained substrates. One class of local learning methods contrasts the desired, clamped behavior with spontaneous, free behavior. However, directly contrasting free and clamped behaviors requires explicit memory. Here, we introduce ‘Temporal Contrastive Learning’, an approach that uses integral feedback in each learning degree of freedom to provide a simple form of implicit non-equilibrium memory. During training, free and clamped behaviors are shown in a sawtooth-like protocol over time. When combined with integral feedback dynamics, these alternating temporal protocols generate an implicit memory necessary for comparing free and clamped behaviors, broadening the range of physical and biological systems capable of contrastive learning. Finally, we show that non-equilibrium dissipation improves learning quality and determine a Landauer-like energy cost of contrastive learning through physical dynamics.