During sleep and quiet wakefulness, the brain spontaneously revisits neural activity patterns associated with previous experience, a phenomenon known as reactivation. Reactivation is considered to support systems consolidation, the process by which memory retrieval becomes progressively more dependent on neocortical circuits and less dependent on the hippocampus. Evidence indicates that reactivation is not indiscriminate but preferentially targets specific memory traces. Systems consolidation has motivated extensive computational modelling of spontaneous reactivation, the prioritisation of specific memories for replay, and the redistribution of memory representations between the hippocampus and neocortex. In parallel, the development of predictive-coding (PC) models has shown how computations attributed to hippocampal and neocortical circuits can arise from local prediction-error dynamics. Here, we connect these two lines of work by showing how autonomous intrinsic dynamics in coupled PC networks can transfer stored memory content between neural systems. The hippocampus and neocortex are modelled as two coupled PC associative memory networks. The architecture incorporates asymmetric coupling motivated by reported interactions between hippocampal area CA1 and the prefrontal cortex. This coupling establishes a closed-loop dialogue in which hippocampal activity trains the neocortical network, while neocortical feedback biases subsequent hippocampal activity toward content for which neocortical support remains weak. The architecture achieved faster transfer of correlated memory content than uniform rehearsal and required no explicit access to the stored memory corpus. Together, these results provide a PC account in which neocortical feedback organises hippocampal reactivation to support efficient memory transfer during systems consolidation.
Blenkmann, A. O., Leske, S. L., Knight, R. et al. · neuroscience
Methods 96%
Abstract
Intracranial EEG (iEEG) affords a unique opportunity to assess human neural activity with millisecond temporal resolution and millimeter spatial precision. However, group-level statistical analyses remain challenging due to sparse and heterogeneous electrode coverage, anatomical variability, and the difficulty of integrating spatiotemporal responses across individuals. Most iEEG studies therefore define brain regions of interest for the group analysis, sacrificing a spatial resolution strength of iEEG and excluding data outside those regions of interest. Here, we present a spatiotemporal analysis and mapping with permutation-based linear mixed-effects models (STAMP-LME), a cortical-mapping framework for population-level iEEG analysis independent of predefined cortical regions of interest. STAMP-LME is a complete processing pipeline spanning electrode localization, mapping to a standardized surface of the cortical mantle, including medial, orbitofrontal, and insular regions, and statistical inference at the cortical level. Electrode-level data from each participant is mapped to a standardized cortical surface using weights that account for electrode-to-cortex distance and confidence in cortical sampling. Brain signals are then combined with a spatial smoothing kernel and entered into LME models at each vertex-node and time point. Data-derived null distributions are generated through controlled permutations and used for spatiotemporal inference with threshold-free cluster enhancement (TFCE) or false-discovery-rate (FDR) correction. We validated STAMP-LME using simulations with known cortical ground truth while varying effect size, interparticipant source-location variability, electrode coverage, and spatial smoothing. The framework reliably detected simulated effects and recovered their spatiotemporal distribution while maintaining low localization error and limiting false-positive detections across different sampling conditions. We then applied the method to intracranial recordings from nine individuals performing an auditory oddball task. The empirical analysis revealed distributed auditory cortical responses consistent with conventional channel-level analyses, while providing more fine-grained group-level activation maps across sampled cortical regions. STAMP-LME provides a flexible, open-source framework for data-driven population-level cortical iEEG analysis that preserves the spatiotemporal strengths of intracranial recordings while accounting for sparse and heterogeneous electrode coverage. The structured workflow enables reproducible and anatomically precise group-level inference on a standardized cortical surface and facilitates comparison with surface-based fMRI and source-localized M/EEG results, therefore providing a practical tool for cognitive and clinical neuroscience research.
Soedirdjo, S., Holobar, A., Dhaher, Y. Y. · neuroscience
Methods 92%
Abstract
Tracking individual motor units (MUs) across multiple testing visits is critical for understanding neuromuscular adaptations and disease progression. However, existing methods fail to account reliably for spatiotemporal variations in motor unit action potentials (MUAPs) and lack validation for longitudinal tracking. Here, we present a robust computational framework for continuous MU tracking that integrates automatic estimation of electrode grid displacements and compensation for MUAP shape variations. Our algorithm demonstrated superior tracking accuracy, capturing 63% and 75% of MUs in the tibialis anterior and medial gastrocnemius, respectively, substantially outperforming state-of-the-art methods. By applying this approach to investigate estradiol-driven neuromuscular plasticity as a model system, we reveal previously undetectable modulations in recruitment thresholds and firing rates, linking hormonal fluctuations to motor unit behavior. Beyond advancing fundamental neuromuscular research, our approach holds promise for enhancing biomarker sensitivity in clinical trials, particularly in conditions such as amyotrophic lateral sclerosis, where reliable MU tracking is paramount. These findings establish a new standard for non-invasive motor unit tracking, unlocking opportunities for longitudinal neurophysiological assessment and evaluation of therapeutics.
Gonzalez-Ferrer, J., Moreno-Ochando, A., Minnick, J. R. et al. · neuroscience
Methods 92%Neural coding 37%
Abstract
Extracellular recordings carry information about neuronal location, and pooling per-unit predictions across neighboring electrodes improves its readout, but whether this local pooling should be learned is untested. We present MYCEL (Message-passing Yields Cellular Embeddings of Location), a graph neural network whose nodes are spike-sorted units carrying waveform and spike-timing features selected on cell-type ground truth, and whose edges encode recording geometry. On the Allen Brain Observatory Visual Coding Neuropixels dataset, learned message passing outperforms fixed aggregation, with the largest gains where single units are least informative. On the International Brain Laboratory Brain Wide Map, MYCEL matches the strongest population model at one-tenth the pairwise interactions. On held-out animals, the embeddings recover laminar depth within visual cortex. Retrained with only the edge geometry changed, the same framework decodes anatomy in macaque visual cortex and human medial temporal lobe. Thus, how neighboring neurons are combined matters as much as the features that describe them.
Neural activity often exhibits multiple timescales that can vary with behavioral states and task conditions. Identifying these timescales from neural recordings is important for better understanding neural computation and function. However, traditional approaches based on autocorrelation fitting are difficult to scale to high-dimensional population recordings and can become unreliable when neural dynamics change with behavior. State-space models have been a powerful framework for modeling high-dimensional neural population activity through latent dynamical systems, but standard formulations and inference methods do not explicitly account for multiple timescales and therefore do not guarantee accurate recovery of the underlying temporal structure. Motivated by these questions, we introduce the Multi-Timescale Switching Linear Dynamical System (MTS-SLDS), a framework for identifying regime-specific latent timescales from continuous or spiking neural observations. MTS-SLDS combines a multi-lag moment initialization, which captures temporal structure across multiple observation lags, with \textit{regime-conditioned} Laplace-EM inference, which reduces mixing of dynamical statistics across uncertain regimes. Characteristic timescales can then be extracted directly from the eigenvalues of the learned latent transition matrices. In synthetic and neural experiments with Gaussian and Poisson spike observations, MTS-SLDS accurately recovers timescales and switching structure over multiple datasets.
Wiafe, S.-L., Soleimani, N., Adali, T. et al. · neuroscience
Methods 91%
Abstract
Blind source separation (BSS) recovers latent sources from observed mixtures. In brain imaging, sources occur in two domains, spatial and temporal, so a criterion applied in one leaves the other unconstrained. Independent component analysis, the most used BSS tool, exploits diversity such as non-Gaussianity and sample dependence, but in one domain at a time. Methods that use both domains apply the same criterion to each and weight them against each other. We propose spatio-dynamical decomposition (SDD), scoring maps by non-Gaussianity and timecourses by dynamical independence: how much each source's dynamics depend on the others. We normalize each score by how much it can improve on the data, eliminating the need for a weight. Across simulations and 130 naturalistic fMRI runs, SDD recovers sources distinguished by either maps or dynamics and reproduces both across disjoint episodes, while baselines trade off the two; map and timescale together also better predict stimulus tracking, with the largest gain for SDD. Together, these results show that separation is strongest when spatial and temporal domains each carry their own criterion.