István Ulbert

Human organotypic slice cultures provide experimental access to adult human neuronal circuits ex vivo, yet it remains unclear whether these networks preserve function or undergo fundamental reorganization following the profound perturbation of slice preparation. Here, a multimodal approach combines extracellular electrophysiology, longitudinal calcium imaging, and quantitative histology to track the changes of human cortical slice cultures over several weeks in vitro. Early phases are marked by pronounced variability and instability, with reduced firing rates, increased burst propensity in principal cells, and heterogeneous recruitment during population activity. These functional changes coincide with substantial structural remodeling, including considerable neuronal loss, disruption of laminar architecture, reactive gliosis, and selective vulnerability of inhibitory interneurons. Despite this structural degradation, neuronal activity converges: electrophysiological properties, cell‐type‐specific firing patterns, and population‐event recruitment become stable and highly consistent. Calcium imaging reveals the progressive evolution of spatially confined active regions, indicating the stabilization of structured network dynamics. These findings demonstrate a remarkable capacity of adult human neuronal circuits for functional adaptation, transitioning from heterogeneous, injury‐driven dynamics to stable and reproducible neuronal activity. This dissociation between anatomical deterioration and functional stabilization should be considered when using organotypic slice cultures as a platform for studying human physiological or pathological brain network dynamics.

Human cortical activity reflects interactions between local recurrent circuits and distributed brain-wide inputs, but how these contributions shape cortical dynamics remains unclear. We compared oscillatory and single unit activity in laminar recordings from the same human cortical regions in eight patients across wakefulness, NREM sleep, and acute slices after surgical isolation. Gamma-band spike-field synchronization increased from wakefulness to sleep and isolated cortex, whereas population coupling, putative connectivity, network integration, and dynamical dimensionality declined. Laminar gamma sink-source organization persisted in isolation, indicating the presence of local gamma-generating mechanisms. Integration with histological data showed preserved tissue architecture, while neuronal density predicted population integration in vivo but not after isolation. Removing long-range inputs thus did not suppress cortical activity but shifted its organization toward stronger rhythmic coordination and reduced integration. Our findings suggest that neocortical microcircuits intrinsically generate coherent gamma activity, whereas network embedding supports the diverse, high-dimensional dynamics of the intact cortex.

Two‐photon calcium imaging offers a powerful way to study neuronal population activity in animal experiments; however, its combination with visual stimulation introduces an inherent technical challenge: light used for stimulation can enter the detection pathway, leading to stray‐light contamination and compromised signal quality. To address this problem, we developed the Photon Shield, an open‐hardware light‐shielding system that combines a 3D‐printed headplate with a rail‐mounted cylindrical shield surrounding the objective. We aimed to create a solution that is fabrication‐accessible and reproducible across laboratories, while being simple to mount and effective in minimizing stimulus‐related artifacts. Quantitative measurements show that the system reduces contaminating photon flux by two orders of magnitude relative to conventional approaches, thereby maintaining fluorescence signal integrity even under high‐intensity visual stimulation. Functional validation in Thy1‐GCaMP6s mice confirms reliable detection of orientation‐selective responses in visual cortex, while demonstrating that the mechanical stability of head fixation is sufficient to support awake, head‐restrained recordings. More broadly, the Photon Shield demonstrates how additive manufacturing introduces a qualitatively new paradigm for experimental tool development. 3D design enables digital specification of often‐overlooked experimental accessories, allowing them to be locally fabricated, improved, and reshared across laboratories, transforming custom solutions into elements of an evolving ecosystem.

Human organotypic slice cultures provide experimental access to adult human neuronal circuits ex vivo, yet it remains unclear whether these networks preserve function or undergo fundamental reorganization following the profound perturbation of slice preparation. Here, we combined extracellular population and single-unit electrophysiology, longitudinal calcium imaging, and quantitative histology to track the changes of human cortical slice cultures over several weeks in vitro. Early phases were marked by pronounced variability and instability, with reduced firing rates, increased burst propensity in principal cells, elevated discharge irregularity, and heterogeneous recruitment during population activity. These functional changes coincided with substantial structural remodeling, including considerable neuronal loss, disruption of laminar architecture, reactive gliosis, and selective vulnerability of inhibitory interneurons. Strikingly, despite this progressive structural degradation, neuronal activity did not diverge but instead converged. By the fourth week in culture, electrophysiological properties, cell-type-specific firing patterns, and population-event recruitment became stable and highly consistent across patients. Calcium imaging revealed persistent, spatially confined regions of synchronous activity, indicating the preservation of structured network dynamics. These events remained within physiological regimes and lacked features of epileptiform discharges. Thus, human neuronal circuits exhibit a robust capacity for self-organization, transitioning from heterogeneous, injury-driven dynamics to stable and homogeneous functional states. This dissociation between structural deterioration and functional convergence establishes human organotypic slice cultures as a reproducible and translationally relevant platform for studying human brain network dynamics ex vivo.

The subthalamic nucleus (STN) is a critical hub for inhibitory control, implicated in decision making under conflict and impulsivity. Delta frequency oscillations have also been associated with inhibitory control processes, yet the relationship between human STN neuronal activity and local delta frequencies during response inhibition remains unresolved. Here we recorded STN neurons and local field potentials in patients with Parkinson's disease performing a stop-signal reaction time task during deep brain stimulation surgery. Approximately half of STN neurons responded to a diverse set of behaviorally relevant events including go and stop signals, with stronger go-related firing and enhanced delta phase coupling linked to failed inhibition. Notably, a specific population of bursting STN neurons showed increased delta coupling. These findings suggest that STN neurons integrate go and stop information, and that enhanced delta engagement and bursting may impair inhibitory control, providing insights into the neuronal mechanism of action cancellation.

Intracortical microstimulation is an essential tool for basic neuroscience and sensory restoration, yet the spatiotemporal effects of advanced multielectrode stimulation strategies on cortical networks remain poorly understood. This study presents a versatile experimental platform featuring polymer-based, flexible multi-shank electrode arrays integrated with a custom high-density neurostimulator. The polyimide penetrating probes contain densely spaced iridium oxide microelectrodes on thin, flexible shanks with sharp tips designed to minimize mechanical mismatch, tissue damage, and brain dimpling. The system was validated in transgenic GCaMP6 mice through simultaneous two-photon calcium imaging and electrical stimulation in layer 2/3 of the visual cortex. Monopolar stimulation reliably evoked robust, parameter-dependent neuronal activation, with neuronal recruitment increasing as a function of current intensity and pulse duration. Bipolar stimulation produced activation patterns distinct from monopolar stimulation, with neuronal responses that were more spatially confined when bipolar electrodes were positioned in close proximity. Furthermore, static current steering - implemented by varying the current ratio between the selected electrode pair - successfully shifted the centroid of activated neuronal populations, enabling fine spatial control over the site of activation. The developed platform provides a flexible framework for investigating local neuronal responses to complex electrical stimulation paradigms, facilitating the development of neural prostheses with higher spatial precision.

Cortical slow waves are hallmark oscillations of deep sleep and certain anesthetic conditions, yet the neurobiological mechanisms controlling their dynamics remain incompletely understood. Here, we investigated the effects of local near-infrared (NIR) stimulation on slow-wave activity in ketamine/xylazine-anesthetized rats. Using a silicon-based multimodal optrode, we simultaneously delivered NIR light and recorded local field potentials (LFPs) and multi-unit activity (MUA) across cortical layers in the primary somatosensory (S1Tr) and parietal association (PtA) cortices. NIR stimulation induces local tissue heating, resulting in reproducible and reversible changes in the properties of slow waves. Specifically, up-state durations were shortened, down-states prolonged, and MUA amplitudes during up-states increased, with steeper slopes at state transitions, indicative of enhanced neuronal synchronization. LFP amplitude and spectral changes varied across cortical regions: PtA exhibited increased slow wave (0.5–2 Hz) and high delta (2–4 Hz) band activity, while S1Tr showed a trend toward reduction. Our findings demonstrate that local infrared stimulation can reliably modulate cortical slow-wave dynamics, likely through temperature-mediated changes in neuronal excitability. This approach provides a minimally invasive method for precise, local manipulation of cortical network activity and offers new insights into the biophysical regulation of slow oscillations.

Cortical slow waves are hallmark oscillations of deep sleep and certain anesthetic conditions, yet the neurobiological mechanisms controlling their dynamics remain incompletely understood. Here, we investigated the effects of local near-infrared (NIR) stimulation on slow-wave activity in ketamine/xylazine-anesthetized rats. Using a silicon-based multimodal optrode, we simultaneously delivered NIR light and recorded local field potentials (LFPs) and multi-unit activity (MUA) across cortical layers in the primary somatosensory (S1Tr) and parietal association (PtA) cortices. NIR stimulation induced local tissue heating, resulting in reproducible and reversible changes in the properties of slow waves. Specifically, up-state durations were shortened, down-states prolonged, and MUA amplitudes during up-states increased, with steeper slopes at state transitions, indicative of enhanced neuronal synchronization. LFP amplitude and spectral changes varied across cortical regions: PtA exhibited increased slow wave (0.5 – 2 Hz) and high delta (2 – 4 Hz) band activity, while S1Tr showed a trend toward reduction. Our findings demonstrate that local infrared stimulation can reliably modulate cortical slow-wave dynamics, likely through temperature-mediated changes in neuronal excitability. This approach provides a minimally invasive method for precise, local manipulation of cortical network activity and offers new insights into the biophysical regulation of slow oscillations.

Recent advancements in high‐density implantable intracortical electrode technology have significantly improved neural interfaces for both research and clinical applications. However, a significant challenge persists: scaling up these devices to achieve recording of nearly all single‐unit activity across large brain volumes. This critical review explores recent progress in neural electrode design, focusing on the challenges of achieving scalable solutions for this ambitious goal. The physical and technical constraints of both rigid and flexible probes are addressed, highlighting the limitations imposed by shank stiffness, mechanical tissue damage, and foreign body response. It is identified that the physics of inserting the electrodes into the brain tissue poses a fundamental constraint, which inherently restricts achievable electrode density. Biohybrid strategies, integrating biological and synthetic components, have shown promise, but they have yet to overcome the major challenges necessary to achieve a scalable functional interface. It is concluded that, given the current limitations of available techniques, there is a pressing need to explore fundamentally novel approaches to realize the vision of recording the electrical activity of every cortical neuron within the brain.

High-density microelectrode arrays have opened new possibilities for systems neuroscience, but brain motion relative to the array poses challenges for downstream analyses. We introduce DREDge (Decentralized Registration of Electrophysiology Data), a robust algorithm for the registration of noisy, nonstationary extracellular electrophysiology recordings. In addition to estimating motion from action potential data, DREDge enables automated, high-temporal-resolution motion tracking in local field potential data. In human intraoperative recordings, DREDge’s local field potential-based tracking reliably recovered evoked potentials and single-unit spike sorting. In recordings of deep probe insertions in nonhuman primates, DREDge tracked motion across centimeters of tissue and several brain regions while mapping single-unit electrophysiological features. DREDge reliably improved motion correction in acute mouse recordings, especially in those made with a recent ultrahigh-density probe. Applying DREDge to recordings from chronic implantations in mice yielded stable motion tracking despite changes in neural activity between experimental sessions. These advances enable automated, scalable registration of electrophysiological data across species, probes and drift types, providing a foundation for downstream analyses of these rich datasets.