Jack J. Lin

Objective . Depressive symptoms are common in individuals with temporal lobe epilepsy (TLE), yet the network dynamics linking limbic circuitry to mood disturbance in TLE remain poorly understood. Here, we investigated whether the temporal organization of amygdala-hippocampal beta synchrony reflects depressive symptom burden in individuals with TLE undergoing intracranial EEG (iEEG) monitoring. Approach . We analyzed iEEG recordings from 14 adults with TLE who underwent intracranial monitoring as part of routine pre-surgical evaluation. Dynamic functional connectivity (dFC) in the beta band was calculated between the amygdala and the hippocampus, and the temporal structure of the dFC was quantified using detrended fluctuation analysis. Random permutation of epochs was performed to assess the influence of interictal epileptiform discharges (IEDs) on the network dynamics. We additionally extracted the ultra-slow fluctuations in dFC and assessed whether the timing of the IEDs was phase-locked to these ultra-slow rhythms. Main results . TLE patients with depression exhibited higher α than non-depressed TLE subjects ( P = 0.007), indicating greater temporal persistence of limbic beta synchrony. Across individuals, α was positively associated with depressive symptom severity ( r = 0.73, P = 0.003). Disrupting the temporal alignment of IED-containing epochs reduced α , whereas permuting IED-containing epochs among themselves preserved α , suggesting that the timing of IEDs contributes to the observed long-range temporal structure. In depressed TLE subjects, IEDs preferentially occurred during the rising phase of ultra-slow dFC fluctuations ( P = 0.004). Significance . Depressive symptoms in TLE are associated with more temporally persistent, less flexible beta band limbic network dynamics. These findings highlight the temporal organization of amygdala-hippocampal synchrony as a potential network-level signature of affective dysfunction within epilepsy and may inform future approaches to monitoring and modulating mood-related circuit activity in this population.

Adaptive behavior relies on the ability to translate abstract rules and goals into actions suited to the current context. Neural population activity in the prefrontal cortex (PFC) has been proposed to support such flexible computations through high-dimensional dynamics, whereas activity in the primary motor cortex (M1) is related more directly to movement execution. How contextual representations in PFC are transformed into ensuing action plans within M1 remains unknown. Previous work suggests that low-dimensional coding subspaces might organize interareal communication, but direct evidence for such population-level communication mechanisms in humans is lacking. Here we use intracranial recordings from human PFC and M1 to identify a communication subspace embedded within high-dimensional PFC activity, that selectively relays behaviorally relevant information at the single-trial level. Activity in this subspace predicts context-dependent action more strongly than either region, revealing a fundamental coding principle by which coordinated interareal population dynamics filter and relay predictive information to guide context-dependent actions.

Choosing to approach or avoid is common in everyday life and excessive avoidance is a cardinal feature of anxiety disorders. We use intracranial EEG to define a prefrontal-limbic circuit supporting approach and avoidance. Presurgical epilepsy patients (n = 20) performed an approach-avoidance conflict decision-making task inspired by the arcade game Pac-Man, where patients trade off rewards against losses from ghost attack. During approach, theta power increases across a limbic circuit including the hippocampus, amygdala, orbitofrontal cortex and anterior cingulate cortex, which drops during avoidance. Theta connectivity between this circuit and lateral prefrontal cortex increases during approach and falls during avoidance. Network connectivity tracks how long patients approach, with enhanced synchronicity extending approach times. During imminent threat, the system switches to sustained increase in high-frequency activity in the lateral prefrontal cortex. The results provide evidence of a distributed prefrontal-limbic circuit, mediated by theta oscillations and high frequency activity, underlying approach-avoidance conflict in humans.

Cognitive flexibility relies on the continuous accumulation and integration of sensory evidence to guide adaptive behavior. In natural environments, behaviorally relevant information unfolds sequentially over time and is constantly evaluated against prior knowledge, task rules, and current demands. Integration of these inputs poses a computational challenge: How is temporally unfolding, predictive information integrated into a stable representation, while preserving the discriminability and flexibility to map individual stimuli to competing context-specific actions? Using large-scale human intracranial electroencephalography, we assessed how neural population activity integrates behaviorally relevant information across multiple sensory events that sequentially unfold over time and jointly determine the current context. The results uncover that the population geometry supports the emergence of conjunctive coding subspaces that integrate prior information with current sensory evidence and jointly define the temporal context that mediates behavioral benefits. Evidence accumulation diversifies the population responses distributed across the cortex, increasing the representational space that embeds context-dependent stimulus-action mappings. Hence, context-dependent sensory coding might constitute the neural basis underlying adaptive human behavior. In sum, these results demonstrate how neural population activity balances integrating predictive information with preserving stimulus discriminability to enable flexibility, while minimizing interference.

Brain-wide communication supporting flexible behavior requires coordination between sensory and associative regions but how brain networks route sensory information at fast timescales to guide action remains unclear. Using human intracranial electrophysiology and spiking neural networks during spatial attention tasks, where participants detected targets at cued locations, we show that high-frequency activity bursts (HFAbs) mark temporal windows of elevated population firing that enable fast, long-range communications. HFAbs were evoked by sensory cues and targets, dynamically coupled to low-frequency rhythms. Notably, both the strength of cue-evoked HFAbs and their decoupling from slow rhythms predicted behavioral accuracy. HFAbs synchronized across the brain, revealing distinct cue- and target-activated subnetworks. These subnetworks exhibited lead–lag dynamics following target onset, with cue-activated subnetworks preceding target-activated subnetworks when cues were informative. Computational modeling suggested that HFAbs reflect transitions to population spiking, denoting temporal windows for network communications supporting attentional performance. These findings establish HFAbs as signatures of population state transitions, supporting information routing across distributed brain networks.

Achieving goals in real-life situations—from fetching a glass of water to landing a dream job—often requires planning based on experience and executing a sequence of actions. Neurophysiological research in animal models has indicated that the orbitofrontal cortex (OFC) mediates relationships between memory, actions, and outcomes and the hippocampus and other medial temporal lobe (MTL) regions are known to be critical for rapid learning, but little is known about how these areas interact to support rapid learning and retrieval of goal-directed action sequences in humans. Here, we leverage a rare opportunity to investigate human OFC gamma oscillations and examine the coordination between the OFC and MTL during a continuous multi-step task that requires applying recently acquired experience to guide behavior. We used multisite intracranial electroencephalography (iEEG) recordings while participants searched for a hidden goal in an animated game to study neural activity in both brain areas during goal-directed behavior. Hippocampal ripples—brief high-frequency oscillations reflecting synchronized neuronal firing—are known to support memory consolidation during sleep, but their role during active memory retrieval and updating remains unclear. We found that OFC gamma activity was modulated by both memory demands and ripples in the hippocampus and adjacent structures. Notably, ripple-coupled OFC gamma during exploration was associated with subsequent task performance. We propose that hippocampal ripples mark a narrow window, supporting hippocampal-cortical communication required for successful goal encoding for future behaviors.

The neurophysiological mechanisms supporting brain maturation are fundamental to attention and memory capacity across the lifespan. Human brain regions develop at different rates, with many regions developing into the third and fourth decades of life. Here, in this preregistered study (https://osf.io/gsru7), we analysed intracranial electroencephalography recordings from widespread brain regions in a large developmental cohort. Using task-based (that is, attention to to-be-remembered visual stimuli) and task-free (resting-state) data from 101 children and adults (5.93–54.00 years, 63 males; n electrodes = 5,691), we mapped aperiodic (1/ƒ-like) activity, a proxy of neural noise, where steeper slopes indicate less noise and flatter slopes indicate more noise. We reveal that aperiodic slopes flatten with age into young adulthood in both association and sensorimotor cortices, challenging models of early sensorimotor development based on brain structure. In the prefrontal cortex (PFC), attentional state modulated age effects, revealing steeper task-based than task-free slopes in adults and the opposite in children, consistent with the development of cognitive control. Age-related differences in task-based slopes also explained age-related gains in memory performance, linking the development of PFC cognitive control to the development of memory. Last, with additional structural imaging measures, we reveal that age-related differences in grey matter volume are similarly associated with aperiodic slopes in association and sensorimotor cortices. Our findings establish developmental trajectories of aperiodic activity in localized brain regions and illuminate the development of PFC control during adolescence in the development of attention and memory.

Background: For patients with drug-resistant focal epilepsy, surgical resection of the epileptogenic zone (EZ) is an effective treatment to control seizures. Accurate localization of the EZ is crucial and is typically achieved through comprehensive presurgical approaches such as seizure semiology interpretation, electroencephalography (EEG), magnetic resonance imaging (MRI), and intracranial EEG (iEEG). However, interpreting seizure semiology is challenging because it heavily relies on expert knowledge. The semiologies are often inconsistent and incoherent, leading to variability and potential limitations in presurgical evaluation. To overcome these challenges, advanced technologies like large language models (LLMs)—with ChatGPT being a notable example—offer valuable tools for analyzing complex textual information, making them well-suited to interpret detailed seizure semiology descriptions and accurately localize the EZ. Objective: This study evaluates the clinical value of ChatGPT for interpreting seizure semiology to localize EZs in presurgical assessments for patients with focal epilepsy and compares its performance with that of epileptologists. Methods: We compiled 2 data cohorts: a publicly sourced cohort of 852 semiology-EZ pairs from 193 peer-reviewed journal publications and a private cohort of 184 semiology-EZ pairs collected from Far Eastern Memorial Hospital (FEMH) in Taiwan. ChatGPT was evaluated to predict the most likely EZ locations using 2 prompt methods: zero-shot prompting (ZSP) and few-shot prompting (FSP). To compare the performance of ChatGPT, 8 epileptologists were recruited to participate in an online survey to interpret 100 randomly selected semiology records. The responses from ChatGPT and epileptologists were compared using 3 metrics: regional sensitivity (RSens), weighted sensitivity (WSens), and net positive inference rate (NPIR). Results: In the publicly sourced cohort, ChatGPT demonstrated high RSens reliability, achieving 80% to 90% for the frontal and temporal lobes; 20% to 40% for the parietal lobe, occipital lobe, and insular cortex; and only 3% for the cingulate cortex. The WSens, which accounts for biased data distribution, consistently exceeded 67%, while the mean NPIR remained around 0. These evaluation results based on the private FEMH cohort are consistent with those from the publicly sourced cohort. A group t test with 1000 bootstrap samples revealed that ChatGPT-4 significantly outperformed epileptologists in RSens for the most frequently implicated EZs, such as the frontal and temporal lobes (P<.001). Additionally, ChatGPT-4 demonstrated superior overall performance in WSens (P<.001). However, no significant differences were observed between ChatGPT and the epileptologists in NPIR, highlighting comparable performance in this metric. Conclusions: ChatGPT demonstrated clinical value as a tool to assist decision-making during epilepsy preoperative workups. With ongoing advancements in LLMs, their reliability and accuracy are anticipated to improve.

Debilitating anxiety is pervasive in the modern world. Choices to approach or avoid are common in everyday life and excessive avoidance is a cardinal feature of anxiety disorders. Here, we used intracranial EEG to define a distributed prefrontal-limbic circuit supporting approach and avoidance. Presurgical epilepsy patients (n=20) performed a continuous-choice, approach-avoidance conflict decision-making task inspired by the arcade game Pac-Man, where patients trade-off harvesting rewards against potential losses from attack by the ghost. As patients approached increasing rewards and threats, we found evidence of a limbic circuit mediated by increased theta power in the hippocampus, amygdala, orbitofrontal cortex (OFC) and anterior cingulate cortex (ACC), that drops rapidly during avoidance. Theta band connectivity within this circuit and with the lateral prefrontal cortex increases during approach and falls during avoidance, and amygdala and lateral frontal activity granger-caused the theta oscillations in both the OFC and ACC. Importantly, the degree of network connectivity predicted how long patients approach, with enhanced network synchronicity extending approach times. Finally, when threat is imminent, the system dynamically switches to a sustained increase in high-frequency activity (70-150Hz) in the middle frontal gyrus (MFG), tracking the degree of threat. The results provide evidence for a distributed prefrontal-limbic circuit, mediated by theta oscillations and high frequency activity, underlying approach-avoidance conflict in humans.

BACKGROUND For patients with drug-resistant focal epilepsy, surgical resection of the epileptogenic zone (EZ) is an effective treatment to control seizures. Accurate localization of the EZ is crucial and is typically achieved through comprehensive presurgical approaches such as seizure semiology interpretation, electroencephalography (EEG), magnetic resonance imaging (MRI), and intracranial EEG (iEEG). However, interpreting seizure semiology is challenging because it heavily relies on expert knowledge. The semiologies are often inconsistent and incoherent, leading to variability and potential limitations in presurgical evaluation. To overcome these challenges, advanced technologies like large language models (LLMs)—with ChatGPT being a notable example—offer valuable tools for analyzing complex textual information, making them well-suited to interpret detailed seizure semiology descriptions and accurately localize the EZ. OBJECTIVE This study evaluates the clinical value of ChatGPT for interpreting seizure semiology to localize EZs in presurgical assessments for patients with focal epilepsy and compares its performance with that of epileptologists. METHODS We compiled 2 data cohorts: a publicly sourced cohort of 852 semiology-EZ pairs from 193 peer-reviewed journal publications and a private cohort of 184 semiology-EZ pairs collected from Far Eastern Memorial Hospital (FEMH) in Taiwan. ChatGPT was evaluated to predict the most likely EZ locations using 2 prompt methods: zero-shot prompting (ZSP) and few-shot prompting (FSP). To compare the performance of ChatGPT, 8 epileptologists were recruited to participate in an online survey to interpret 100 randomly selected semiology records. The responses from ChatGPT and epileptologists were compared using 3 metrics: regional sensitivity (RSens), weighted sensitivity (WSens), and net positive inference rate (NPIR). RESULTS In the publicly sourced cohort, ChatGPT demonstrated high RSens reliability, achieving 80% to 90% for the frontal and temporal lobes; 20% to 40% for the parietal lobe, occipital lobe, and insular cortex; and only 3% for the cingulate cortex. The WSens, which accounts for biased data distribution, consistently exceeded 67%, while the mean NPIR remained around 0. These evaluation results based on the private FEMH cohort are consistent with those from the publicly sourced cohort. A group t test with 1000 bootstrap samples revealed that ChatGPT-4 significantly outperformed epileptologists in RSens for the most frequently implicated EZs, such as the frontal and temporal lobes (P<.001). Additionally, ChatGPT-4 demonstrated superior overall performance in WSens (P<.001). However, no significant differences were observed between ChatGPT and the epileptologists in NPIR, highlighting comparable performance in this metric. CONCLUSIONS ChatGPT demonstrated clinical value as a tool to assist decision-making during epilepsy preoperative workups. With ongoing advancements in LLMs, their reliability and accuracy are anticipated to improve.