Biology subjects

Davis, T. S.

Publications and source records attributed to Davis, T. S..

12 recordsLinked to original sources

bioRxiv2026

Low-dimensional factorized neural computations underlie risk-adaptive choices

Real-world decision-making rarely occurs with perfect information. Instead, individuals must constantly weigh potential rewards against the probability of adverse outcomes.1 Failures of this process can lead to maladaptive decisions associated with reduced lifetime success, and numerous psychiatric disorders such as gambling addictions, bulimia nervosa, and substance use disorder.2,3 The neural computations that facilitate inference about the landscape of potential outcomes remain unclear, but are thought to occur in distributed frontotemporal circuits.4 Here we used deep reinforcement learning agents to predict distinct behavioral strategies and their underlying neural population dynamics during a risky decision-making task. Across a range of training conditions, deep reinforcement learning agents separated into strategies marked by either overly cautious exploration of the reward contingency space or a high-performing, risk-adaptive Bimodal strategy. The internal dynamics of high-performing Bimodal agents formed low-dimensional representations that segregated safe and risky states. In contrast, the cautious exploration agents were associated with more skewed and entangled neural representations. We found remarkably similar dynamical representations and their associated behavioral strategies in neuronal ensemble recordings from human epilepsy patients performing a similar risky decision-making task. These results reveal the structure of dynamical computations that underlie inferences about uncertain outcomes and their associated behavioral strategies.

neuroscience↗

bioRxiv2026

Temporal Processing during Decision Making under Uncertainty.

Time perception and decision-making are interrelated processes that are critical to much of human and animal behavior. Although the behavioral importance of this link between the processes has been established, the neural underpinnings of temporal decision-making are unclear. The current study leverages human intracranial electro-physiology to investigate the neural contribution of temporal processing to decision-making under risk. To probe temporal decision-making, we use a version of the Balloon Analogue Risk Task, in which participants inflate virtual balloons with the goal of reaching maximal balloon size but stopping inflation before the balloon pops. Points are awarded based on inflation duration of successful trials, such that points awarded are linearly related to balloon size. By decoding temporal features of the task on a trial-by-trial basis, we find neural representations of time during balloon inflation. Specifically, we find that both medial and lateral temporal regions of the brain encode the majority of this time-related signal. Importantly, we can decode the progression of time leading up to the outcome of a successful decision (pressing a button to stop inflation before the balloon pops). This encoding of time depends on an active decision to end the trial: we cannot decode time during passive observation of balloon inflation to a known maximal size. To validate these results further, we use a temporal convolution network to predict participants response (i.e., time of button press) based on the brain activity leading up to the response. We find that we can predict the moment of response with high accuracy and temporal resolution. The model predicts the timing of successful decisions but not the timing of balloon pops. Together, these findings reveal neural representations that encode the progression of time, specifically in relation to an active decision.

neuroscience↗

bioRxiv2026

Propofol-induced loss of responsiveness reorganizes cortical traveling waves in the human brain

Consciousness depends on the coordinated propagation of neural activity across cortical networks. Traveling waves, the spatiotemporal patterns of phase-aligned neural activity, are thought to support large-scale cortical communication, yet how these dynamics are altered during general anesthesia remains poorly understood. Here, we examined the effects of propofol on cortical traveling waves recorded with high-density microelectrode arrays in the temporal lobes of two human participants (both men). We identified and characterized changes in neural traveling wave properties as participants underwent general anesthesia. Propofol produced robust, state-dependent reorganization of wave dynamics: increased propagation speed, shifted propagation directions, and altered spectral structure. At the neuronal level, propofol was accompanied by pronounced changes in firing activity and spike-wave relationships, linking mesoscale traveling waves to coordinated changes in neuronal firing dynamics. Together, these findings demonstrate how propofol reshapes the spatiotemporal organization of cortical activity across spatial scales.

neuroscience↗

bioRxiv2026

Asymmetric Reinforcement Learning Explains Human Choice Patterns in Decision-making Under Risk

Human decisions under uncertainty are shaped by experience, but the computations that translate expectation and experience into choice remain debated in neural and cognitive science. Prior studies highlight reinforcement learning (RL) as a unifying framework, yet it is unclear whether human behavior under risk is better captured by symmetric updating from outcomes or by asymmetric learning that weights reward and loss differently. This work examines which learning strategies better explain trial-by-trial choices given contextual uncertainty and manipulations of outcome distributions. Our results show that a Risk Sensitive (RS) model with asymmetric learning rates best explains human behavior in our novel decision-making task. Fitting candidate models to individual trial histories yielded value signals that predicted both choice and response time. These results highlight that RS model, as an asymmetric learning provides a concise and identifiable account of behavior in decision-making under risk tasks.

neuroscience↗

bioRxiv2025

Increased Aperiodic Exponents Track Depression Symptom Severity

Roughly one-third of patients with major depressive disorder (MDD) fail to respond to standard treatments and develop treatment-resistant MDD. For these patients, alternative therapies ofer additional options but yield inconsistent outcomes. Progress has been limited by the absence of objective, brain-based biomarkers to guide target selection or track therapeutic response in real time. Instead, clinicians rely on behavioral assessments that evolve slowly over weeks to months, obscuring the underlying neural dynamics of symptom changes. Here, we test whether the aperiodic exponent of intracranial EEG (iEEG) local field potentials can serve as a neurophysiological marker of depression symptom severity. We leveraged a large iEEG cohort (N = 20) undergoing invasive monitoring for refractory epilepsy, yielding over 1,800 contacts spanning cortical and subcortical zones. For each contact, we estimated the aperiodic exponent (thought to reflect aspects of cortical excitability) of the power spectrum across 10-100 Hz within local brain regions and across distributed cortical association networks. Depressive symptoms were assessed with the Beck Depression Inventory-II (BDI-II) immediately before intracranial resting state recordings. With respect to the BDI-II scale, participants were identified as experiencing minimal (BDI-II [≤] 13) or elevated depression symptoms (BDI-II [≥] 14). Associations between symptom severity (BDI-II total score and Somatic-Afective, Cognitive, and Anhedonia subscales) and region- or network-level exponents were modeled with ordinary least squares (OLS) regression. The whole-brain, mean aperiodic exponent for each participant discriminated symptom status (AUC = 0.82). At the regional level, the orbitofrontal cortex, anterior cingulate cortex, insula, and amygdala showed higher exponents in the elevated depression symptom group (d = 1.18-1.71; p = 0.032-0.004). A post-hoc classification analysis across these four regions misclassified one participant per group (AUC = 0.86; 95% CI 0.64-1.00). In continuous analyses, BDI-II scores correlated positively with exponents in these same four regions (pFDR = 0.019-0.027; partial r=0.61-0.70) and at the network level in the Salience network (pFDR = 0.024; partial r = 0.63) and Default (pFDR = 0.046; partial r = 0.55) network. The Salience network significantly tracked Anhedonia symptoms (p = 0.004; partial r = 0.62). Here we report that intracranial aperiodic exponents within fronto-limbic and insular circuits, overlapping with networks implicated in contemporary accounts of depression pathophysiology, diferentiate depressive symptom status and scale with severity. These findings support the aperiodic exponent as a candidate neurophysiological marker of current depression symptom burden, with potential relevance for individualized neuromodulation in MDD.

neuroscience↗

bioRxiv2025

Frontotemporal bursting supports human working memory

Cortical neural activity varies dynamically during memory periods, when relevant information is not present in the environment. But how those dynamics are related to a code defining working memory (WM) performance is not known. Recent data shows brief bursts of activity in the high gamma (70-140 Hz) and beta (12-30 Hz) band within non-human primate lateral prefrontal cortex (PFC) is associated with WM processing. However, WM may be related to activity within a network of frontal executive and posterior sensory areas involved in stimulus perception. Here we tested whether gamma and beta bursting exist in lateral PFC and multisensory lateral temporal areas in humans during visual WM, and whether these areas are coupled via a phase-burst code. We used intracranial macroelectrode recordings from the middle frontal gyrus (MFG), which includes dorsolateral PFC, and from the middle temporal gyrus (MTG), an area important for visual processing. High gamma bursting increased in human left PFC during encoding and delay periods while beta bursting decreased. Interestingly, beta bursting increased in multisensory areas during encoding and remained high during the delay period, more so on the right. These effects varied with WM performance. Finally, we quantify the degree to which delay-period gamma bursting is locked to beta phase within and between regions of this network using a proposed metric termed phase-burst coupling (PBC). We find evidence that delay-period gamma bursting in temporal areas is locked to beta phase in PFC. Our findings suggest that WM may use bursting to support memory maintenance until readout. Significance StatementWorking memory is critical for everyday functioning and its disruption results in significant disability in patients with dementia, traumatic brain injury, and stroke. We recorded human intracranial data during a visual working memory task. We found that prefrontal cortex and multisensory temporal brain areas show neural bursting activity during the memory period and that such bursting is related to trial performance. Furthermore, these two areas may be coupled by a phase-burst code. Bursting has been found in non-human primates, rodents, and even crows, suggesting it is a general feature of memory, and perhaps of other aspects of cognition.

neuroscience↗

bioRxiv2024

Cell-type specific responses to single-pulse electrical stimulation of the human brain

Neuromodulation techniques, such as deep brain stimulation, intraoperative brain mapping, and responsive neurostimulation, use electricity to alter brain activity. Despite daily clinical use in thousands of patients, it remains fundamentally unknown how human neurons respond to intracranial stimulation. We address this question at a basic level by characterizing neuronal cell-type specific firing rate responses to single pulses of electrical stimulation of the human brain. We carried out broadly distributed stimulation in 30 patients undergoing neuromonitoring for epilepsy while recording from isolated neurons on microwires implanted into the medial temporal and frontal lobes. Out of a total of 228 recorded units, 16.2% (N = 191) were classified as interneurons and 83.8% (N = 37) were classified as principal cells, using a threshold clustering method, based on intrinsic waveshape characteristics. To see how stimulation affected neuronal activation for each cell type, we calculated firing rate change between a pre-stimulation and post-stimulation window and observed that 174 units were significantly modulated with the vast majority (91%) showing firing rate suppression. We then characterized stimulation-evoked changes in firing rate to gain insight into cell type-specific responses. Additionally, in a subset of the units, we observed that firing rate responses were modulated by stimulation distance, where local stimulation (within approximately 40 mm) could evoke instantaneous firing, whereas distant stimulation reliably suppressed firing in the same units. Finally, we analyzed a subset of units within the seizure onset zone, which exhibited unique waveform features and responses to stimulation. This study bridges a gap in the neuromodulation field by examining the single-unit firing rate response to direct electrical stimulation of the human brain and analyzing cell-type specific firing rate responses. We show that low frequency, single-pulse stimulation broadly elicits firing rate suppression, but parameters, such as distance from the unit, can have diverse effects on firing rate responses. This work informs the neuronal basis of CCEP generation and therefore has implications for clinical mapping and informs novel active probing strategies for precision diagnosis and neuromodulation of seizure pathophysiology in surgical cases. Moreover, this research has general implications for understanding neuromodulation via direct brain stimulation. Highlights- Putative principal cell waveform shapes are characterized by longer trough-to-peak and full-width half max durations (s) compared to interneurons. - Monopolar stimulation @ 3 mA generally has a widespread suppressive effect on neuronal firing, lasting approximately 1.5 s. - Principal cells show greater suppression amplitude and longer suppression durations than interneurons. - Stimulation within [~] 40 mm is capable of evoking instantaneous firing, despite general suppression from stimulation at greater distances.

neuroscience↗

bioRxiv2023

Subsets of cortico-cortical evoked potentials propagate as traveling waves

Understanding the spatiotemporal dynamics of neural signal propagation is fundamental to unraveling the complexities of brain function. Emerging evidence suggests that cortico-cortical evoked potentials (CCEPs) resulting from single-pulse electrical stimulation may be used to characterize the patterns of information flow between and within brain networks. At present, the basic spatiotemporal dynamics of CCEP propagation cortically and subcortically are incompletely understood. We hypothesized that single-pulse electrical stimulation evokes neural traveling waves detectable in the three-dimensional space sampled by intracranial stereoelectroencephalography. Across a cohort of 21 adult patients with intractable epilepsy, we delivered 17,631 stimulation pulses and recorded CCEP responses in 1,019 electrode contacts. The distance between each pair of electrode contacts was approximated using three different metrics (Euclidean distance, path length, and geodesic distance), representing direct, tractographic, and transcortical propagation, respectively. For each robust CCEP, we extracted amplitude-, spectral-, and phase-based features to identify traveling waves emanating from the site of stimulation. Many evoked responses to stimulation appear to propagate as traveling waves ([~]14-28%), despite sparse sampling throughout the brain. These stimulation-evoked traveling waves exhibited biologically plausible propagation velocities (range 0.1-9.6 m/s). Our results reveal that direct electrical stimulation elicits neural activity with variable spatiotemporal dynamics, including the initiation of neural traveling waves. Significance StatementUsing single-pulse stimulation, we identify a subset of intracranial evoked potentials that propagate as neural traveling waves. Our results were robust across a range of distinct but complementary analysis methods. The identification of stimulation-evoked traveling waves may help to better characterize the pathways traversed by spontaneous, pathological, or task-evoked traveling waves and distinguish biologically plausible propagation from volume-conducted signals.

neuroscience↗

bioRxiv2023

Characterization of spatiotemporal dynamics of binary and graded tonic pain in humans using intracranial recordings

Pain is a complex experience involving sensory, emotional, and cognitive aspects, and multiple networks manage its processing in the brain. Examining how pain transforms into a behavioral response can shed light on the networks relationships and facilitate interventions to treat chronic pain. However, studies using high spatial and temporal resolution methods to investigate the neural encoding of pain and its psychophysical correlates have been limited. We recorded from intracranial stereo-EEG (sEEG) electrodes implanted in sixteen different brain regions of twenty patients who underwent psychophysical pain testing consisting of a tonic thermal stimulus to the hand. Broadband high-frequency local field potential amplitude (HFA; 70-150 Hz) was isolated to investigate the relationship between the ongoing neural activity and the resulting psychophysical pain evaluations. Two different generalized linear mixed-effects models (GLME) were employed to assess the neural representations underlying binary and graded pain psychophysics. The first model examined the relationship between HFA and whether the patient responded "yes" or "no" to whether the trial was painful. The second model investigated the relationship between HFA and how painful the stimulus was rated on a visual analog scale. GLMEs revealed that HFA in the inferior temporal gyrus (ITG), superior frontal gyrus (SFG), and superior temporal gyrus (STG) predicted painful responses at stimulus onset. An increase in HFA in the orbitofrontal cortex (OFC), SFG, and striatum predicted pain responses at stimulus offset. Numerous regions including the anterior cingulate cortex, hippocampus, IFG, MTG, OFC, and striatum, predicted the pain rating at stimulus onset. However, only the amygdala and fusiform gyrus predicted increased pain ratings at stimulus offset. We characterized the spatiotemporal representations of binary and graded painful responses during tonic pain stimuli. Our study provides evidence from intracranial recordings that the neural encoding of psychophysical pain changes over time during a tonic thermal stimulus, with different brain regions being predictive of pain at the beginning and end of the stimulus. Significance StatementWe investigated the neural encoding of pain psychophysics across 16 brain regions during a continuous thermal stimulus in humans. Mixed-effects models were used to analyze trends across 20 human subjects. Using intracranial electrodes, we show a parametric relationship between behavioral responses and HFA during ongoing pain. We found that HFA in cognitive and emotional pain processing regions was closely associated with pain evaluation at the stimulus onset, end, or both. The neural encoding of subjective pain intensity, measured by a visual analog scale, differed from that of binary pain intensity. Perception and psychophysical correlates to pain depend on how patients are asked to evaluate it. Our findings provide evidence that HFA can serve as a neural marker within specific brain regions of behavioral pain responses, as measured by sEEG.

neuroscience↗

bioRxiv2021

Human interictal epileptiform discharges are traveling waves reflecting ictal self-organization

Interictal epileptiform discharges (IEDs), also known as interictal spikes, are large intermittent electrophysiological events observed between seizures in patients with epilepsy. Though they occur far more often than seizures, IEDs are less studied, and their relationship to seizures remains unclear. To better understand this relationship, we examined multi-day recordings of microelectrode arrays implanted in human epilepsy patients, allowing us to precisely observe the spatiotemporal propagation of IEDs, spontaneous seizures, and how they relate. These recordings showed that the majority of IEDs are traveling waves, traversing the same path as ictal discharges during seizures, and with a fixed direction relative to seizure propagation. Moreover, the majority of IEDs, like ictal discharges, were bidirectional, with one predominant and a second, less frequent antipodal direction. These results reveal a fundamental spatiotemporal similarity between IEDs and ictal discharges. These results also imply that most IEDs arise in brain tissue outside the site of seizure onset and propagate toward it, indicating that the propagation of IEDs provides useful information for localizing the seizure focus.

neuroscience↗

bioRxiv2020

Livestock grazing is associated with seasonal reduction in pollinator biodiversity and functional dispersion but cheatgrass invasion is not: variation in bee assemblages in a multi-use shortgrass prairie

Livestock grazing and non-native plant species affect rangeland habitats globally. These factors may have important effects on ecosystem services including pollination, yet, interactions between pollinators, grazing, and invasive plants are poorly understood. To address this, we tested the hypothesis that cattle grazing and site colonization by cheatgrass (Bromus tectorum) impact bee foraging and nesting habitats, and the biodiversity of wild bee communities, in a shortgrass prairie system. Bee nesting habitats (litter and wood cover) were marginally improved in non-grazed sites, though foraging habitat (floral cover and richness) did not differ among grazed, non-grazed, or cheatgrass colonized sites. However, floral cover was a good predictor of bee abundance and functional dispersion. Mean bee abundance, richness, diversity and functional diversity were significantly lower in cattle-grazed habitats than in cheatgrass-colonized or non-grazed habitats. Differences in bee diversity among habitats were pronounced early in the growing season (May) but by late-season (August) these differences eroded. Fourth-corner analysis revealed that sites with high floral cover tended to support large, social, polylectic bees; sites with high grass cover tended to support oligolectic solitary bees. Both cattle-grazed and cheatgrass-colonized sites were associated with lower abundances of above-ground nesting bees but higher abundance of below-ground nesters. We conclude that cheatgrass-invaded sites are not associated with reduced bee biodiversity or abundance, but cattle grazing was negatively associated with bee abundances and altered species composition. Although floral cover is an important predictor of bee assemblages, this was not impacted by grazing and our suggests that cattle likely impact bee communities through effects other than those mediated by forbs, including soil disturbance or nest destruction. Efforts aimed at pollinator conservation in prairie habitats should focus on managing cattle impacts early in the growing season to benefit sensitive bee species.

ecology↗

bioRxiv2020

Long-Term Performance of Utah Slanted Electrode Arrays and Intramuscular Electromyographic Leads Implanted Chronically in Human Arm Nerves and Muscles

ObjectiveWe explore the long-term performance and stability of seven percutanous Utah Slanted Electrode Arrays (USEAs) and intramuscular recording leads (iEMGs) implanted chronically in the residual arm nerves and muscles of three human amputees as a means to permanently restore sensorimotor function after upper-limb. ApproachWe quantify the number of functional recording and functional stimulating electrodes over time. We also calculate the signal-to-noise ratio of USEA and iEMG recordings and quantify the stimulation amplitude necessary to evoke detectable sensory percepts. Furthermore, we quantify the consistency of the sensory modality, receptive field location, and receptive field size of USEA-evoked percepts. Main ResultsIn the most recent subject, involving USEAs with technical improvements, neural recordings persisted for 502 days (entire implant duration) and the number of functional recording electrodes for one USEA increased over time. However, for six out of seven USEAs the number of functional recording electrodes decreased within the first two months after implantation. The signal-to-noise ratio of neural recordings and electromyographic recordings stayed relatively consistent over time. Sensory percepts were consistently evoked over the span of 14 months, were not significantly different in size, and highlighted the nerves fascicular organization. The percentage of percepts with consistent modality or consistent receptive field location between sessions (~1 month apart) varied between 0-86.2% and 9.1-100%, respectively. Stimulation thresholds and electrode impedances increased initially but then remained relatively stable over time. SignificanceThis work demonstrates improved performance of USEAs, and provides a basis for comparing the longevity and stability of USEAs to that of other neural interfaces. Although USEAs provide a rich repertoire of neural recordings and sensory percepts, performance still generally declines over time. Future work should leverage the results presented here to further improve USEA design or to develop adaptive algorithms that can maintain a high level of performance.

neuroscience↗