Biological organisms have sensors that communicate information about the environment. Analyzing how well these biological sensors function has usually been done with mutual information between the sensor signal and the environment, but that can be computationally intractable and summarize something quite complex with just a single number. We suggest that alternatively, one may profitably analyze these biosensors using bias and variance or confusion matrices, depending on the kind of environment. Stimulus-dependent Maximum Entropy models are used to develop estimators of the environmental state given the sensor state, and these estimators in turn are then used to calculate either the bias and variance of the estimator or confusion matrices. We focus on several examples to understand the utility of non-information-based analyses: ligand-receptor binding models spanning genetic regulation to neuronal communication to bacterial chemotaxis, and spin-glass Ising models for neural activity in cultured neurons. These new computationally-efficient analyses add insight to existing analyses based on mutual information; in particular, mutual information estimates give one number to characterize responses to all environmental inputs, and this analysis method characterizes how sensors respond to each environmental input. Categorical analyses, meanwhile, indicate the presence of memory without much prediction in confusion matrix elements in cultured neural networks, adding to previous understanding from mutual information estimates. Author summary All living organisms use external stimuli to navigate their environment via their sensors. Because encoding information costs energy, organisms retain only a fraction of the information received from their sensors, ideally information that maximizes their ability to remember past environmental states or predict future ones, key functions that support survival. To better understand how well sensor systems absorb stimulus information, we used stimulus-dependent Maximum Entropy (MaxEnt) models with maximum likelihood estimation and typical statistical metrics, such as confusion matrices or bias and variance. This approach provides two primary benefits over previous approaches: it is more computationally efficient, and it provides a more information-rich picture on how sensors interact with stimulus. ### Competing Interest Statement The authors have declared no competing interest.

Biological organisms have sensors that communicate information about the environment. Analyzing how well these biological sensors function has usually been done with mutual information between the sensor signal and the environment, but that can be computationally intractable and summarize something quite complex with just a single number. We suggest that alternatively, one may profitably analyze these biosensors using bias and variance or confusion matrices, depending on the kind of environment. Stimulus-dependent Maximum Entropy models are used to develop estimators of the environmental state given the sensor state, and these estimators in turn are then used to calculate either the bias and variance of the estimator or confusion matrices. We focus on several examples to understand the utility of non-information-based analyses: ligand-receptor binding models spanning from genetic regulation to neuronal communication to bacterial chemotaxis and spin-glass Ising models for neural activity in cultured neurons. These new computationally efficient analyses add insight into existing analyses based on mutual information; in particular, mutual information estimates give one number to characterize responses to all environmental inputs, and this analysis method characterizes how sensors respond to each environmental input. Categorical analyses, meanwhile, indicate the presence of memory without much prediction in confusion matrix elements in cultured neural networks, adding to previous understanding from mutual information estimates.

Objective. Memory has been extensively studied at the behavioural as well as the cellular level. Spike timing dependent plasticity is widely considered essential for long-term memory and is associated with activation of N-methyl-D-aspartate (NMDA) receptors. This suggests that NMDA receptor activation plays a crucial role in enabling long-term memory. However, experimental evidence remains sparse, probably due to the complex combination of cellular and functional readouts required. Approach. Recent work showed that in-vitro cortical networks memorize and predict inputs. The initial dependency of prediction on short-term memory decreased during the formation of long-term memory traces. Here, we stimulated networks of dissociated cortical neurons that were grown on multi electrode arrays to investigate memory and prediction under control conditions, or under NMDA block. Main results. The NMDA antagonist 2-amino-5-phosphonovaleric acid (APV) at the used concentration impeded long-term memory trace formation, but did not significantly reduce network excitability. In APV-treated cultures short-term memory of stimuli persisted and they were still able to predict. In contrast to control cultures, prediction remained fully dependent on short-term memory. Significance. This confirms that NMDA receptor activation is essential for the formation of long-term memory traces and supports the notion that, as control cultures learn to memorize the stimulus, long-term memory starts to contribute to their predictive capability.

Low activity in the ischemic penumbra has been suggested as a critical factor leading to apoptosis. BDNF excretion is activity-dependent and has been shown to be neuroprotective. Blockade of TrkB receptors induced apoptosis in cultured neurons, comparable with the effects of hypoxia. BDNF administration was effective in mitigating hypoxia-induced apoptosis, suggesting that the beneficial effect of mild activation to prevent apoptosis in hypoxic cultures is mediated by BDNF-TrkB signaling.

Memory consolidation involves repeated replay of new information by the hippocampus, which transfers memories to the neocortex for long-term storage. This occurs mainly during slow wave sleep, a phase characterized in the cortex by low cholinergic tone and low afferent input. High cholinergic tone has been shown to hamper memory consolidation, probably mediated by reduced network excitability (the ease of activity propagation in a network). We used cortical neuronal networks on multi electrode arrays to investigate whether low background input contributes to memory consolidation. Networks received focal electrical stimuli to memorize, with or without background afferent input (global optogenetic stimulation). Background stimulation hampered memory formation and consolidation, confirming the importance of low background input. Moreover, it lowered network excitability, similar to high cholinergic tone. These findings suggest that high network excitability is a critical feature of slow wave sleep that facilitates memory consolidation.

Importance Out-of-hospital cardiac arrest survival rates have markedly risen in the last decades, but neurological outcome only improved marginally. Despite research on more than 20 neuroprotective strategies involving patients in comas after cardiac arrest, none have demonstrated unequivocal evidence of efficacy; however, treatment with acyl-ghrelin has shown improved functional and histological brain recovery in experimental models of cardiac arrest and was safe in a wide variety of human study populations. Objective To determine safety and potential efficacy of intravenous acyl-ghrelin to improve neurological outcome in patients in a coma after cardiac arrest. Design, Setting, and Participants A phase 2, double-blind, placebo-controlled, multicenter, randomized clinical trial, Ghrelin Treatment of Comatose Patients After Cardiac Arrest: A Clinical Trial to Promote Cerebral Recovery (GRECO), was conducted between January 18, 2019, and October 17, 2022. Adult patients 18 years or older who were in a comatose state after cardiac arrest were assessed for eligibility; patients were from 3 intensive care units in the Netherlands. Expected death within 48 hours or unfeasibility of treatment initiation within 12 hours were exclusion criteria. Interventions Patients were randomized to receive intravenous acyl-ghrelin, 600 mu g (intervention group), or placebo (control group) within 12 hours after cardiac arrest, continued for 7 days, twice daily, in addition to standard care. Main Outcomes and Measures Primary outcome was the score on the Cerebral Performance Categories (CPC) scale at 6 months. Safety outcomes included any serious adverse events. Secondary outcomes were mortality and neuron-specific enolase (NSE) levels on days 1 and 3. Results A total of 783 adult patients in a coma after cardiac arrest were assessed for eligibility, and 160 patients (median [IQR] age, 68 [57-75] years; 120 male [75%]) were enrolled. A total of 81 patients (51%) were assigned to the intervention group, and 79 (49%) were assigned to the control group. The common odds ratio (OR) for any CPC improvement in the intervention group was 1.78 (95% CI, 0.98-3.22; P = .06). This was consistent over all CPC categories. Mean (SD) NSE levels on day 1 after cardiac arrest were significantly lower in the intervention group (34 [6] mu g/L vs 56 [13] mu g/L; P = .04) and on day 3 (28 [6] mu g/L vs 52 [14] mu g/L; P = .08). Serious adverse events were comparable in incidence and type between the groups. Mortality was 37% (30 of 81) in the intervention group vs 51% (40 of 79) in the control group (absolute risk reduction, 14%; 95% CI, -2% to 29%; P = .08). Conclusions and Relevance In patients in a coma after cardiac arrest, intravenous treatment with acyl-ghrelin was safe and potentially effective to improve neurological outcome. Phase 3 trials are needed for conclusive evidence. Trial Registration Clinicaltrialsregister.eu: EUCTR2018-000005-23-NL

INTRODUCTION:Hyperglycemia is common in acute ischemic stroke, and associated with unfavorable outcome. However, the optimal glucose level is not known and cellular effects of hyperglycemia under hypoxia are largely unclear. We assessed how the extracellular glucose concentration affects cultured neuronal networks under experimental in vitro conditions, to provide a starting point for assessment of mechanisms at the neuronal network and cellular levels. METHODS:We used in vitro cultured rat neuronal networks on micro-electrode arrays (MEAs) and glass coverslips. Twenty-four hours of controlled hypoxia was induced. We measured neuronal network activity during baseline (normoxia, 6 h), 24 h of hypoxia, and 6 h after reoxygenation, defined as the summed number of action potentials in 1 h bins. Apoptosis was determined intermittently with caspase 3/7 staining and microscopy. We compared groups of networks under glucose concentrations of 5.0 mmol/L, 7.0 mmol/L, 9.0 mmol/L, and 12.0 mmol/L. RESULTS:Overall, during hypoxia, a gradual decrease in neuronal network activity and increase in apoptosis was found. There were faster decrease in activity (p < 0.01) and more apoptosis after 24 h of hypoxia under glucose levels of 12 mmol/L in a single-well MEA set-up (p < 0.05), and more apoptosis in glass coverslips with glucose levels of 12.0 mmol/L in comparison with 5 mmol/L (p = 0.03). These differences were not observed in multi-well MEAs, in which effects of hypoxia were much smaller than in single-well MEAs. CONCLUSION:Hyperglycemia was associated with a more rapid decrease of neuronal network activity during and more apoptosis after 24 h of hypoxia in cultured neuronal networks. Loss of neuronal activity and apoptosis probably play a role in poorer outcomes of stroke patients under hyperglycemia. Our model provides a starting point for further assessment of pathomechanisms.

Memory has been extensively studied at the behavioural as well as the cellular level. Spike timing dependent plasticity (STDP) is associated with N-methyl-D-aspartate (NMDA) receptor activation and is widely accepted to be essential for long-term memory. However, experimental evidence remains sparse, probably due to the required complex combination of cellular and functional readouts. Recent work showed that in-vitro cortical networks memorize and predict inputs. The initial dependency of prediction on short-term memory decreased during the formation of long-term memory traces. Here, we stimulated in-vitro networks to investigate memory and prediction under control conditions, or under NMDA block. The NMDA anatagonist 2-amino-5-phosphonovaleric acid (APV) at a concentration that did not significantly reduce network excitability, but did impede long-term memory trace formation. In APV-treated cultures short-term memory of stimuli persisted, and they were still able to predict. In contrast to control cultures, prediction remained fully dependent on short-term memory. This confirms that NMDA receptor activation is essential for the formation of long-term memory traces and supports the notion that, as control cultures learn to memorize the stimulus, long-term memory starts to contribute to their predictive capability.### Competing Interest StatementThe authors have declared no competing interest.

Aims: Effective treatments to improve brain recovery after cardiac arrest are needed. Ghrelin showed efficacy in experimental models and was associated with lower neuron-specific enolase levels in the clinical Ghrelin in Coma (GRECO) trial. Here, we present cognitive and psychosocial outcomes at 1-year follow-up. Methods and results: GRECO was a Phase 2 multicentre, double-blind, randomized, placebo-controlled trial in comatose patients after cardiac arrest. The intervention was intravenous acyl-ghrelin 600 mu g twice daily or placebo for 1 week, starting within 12 h after the arrest. Patients were assessed after 1 year using cognitive tests and questionnaires measuring participation, health-related quality of life, mood, and caregiver strain. Composite z-scores of the cognitive tests were computed by comparing the scores with those of a norm population and averaging the tests for memory, attention, and executive functioning separately. Groups were compared based on composite z-scores and cut-off scores for psychosocial outcomes. Of the 160 participants originally included, 66 of the 85 participants who survived to 1 year after OHCA completed the psychosocial and cognitive follow-up. The intervention group scored numerically higher across the cognitive domains compared with the control group, but the differences were not statistically significant (memory median = -0.850 vs. -1.385, U = 424.5, P = 0.587; attention median = -0.733 vs. -0.717, U = 420.5, P = 0.548; and executive functioning median = -0.311 vs. -0.482, U = 408.5, P = 0.323). There were significantly fewer signs of depression in the intervention group (U = 322.5, P = 0.014). Conclusion: This predefined secondary analysis found that ghrelin treatment was associated with non-significantly but consistently better cognitive outcomes and significantly fewer signs of depression. This is in line with the primary outcomes.

Introduction In the core of a brain infarct, perfusion is severely impeded, and neuronal death occurs within minutes. In the penumbra, an area near the core with more remaining perfusion, cells initially remain viable, but activity is significantly reduced. In principle, the penumbra can be saved if reperfusion is established on time, making it a promising target for treatment. In vitro models with cultured neurons on microelectrode arrays (MEAs) provide a useful tool to investigate how ischemic stroke affects neuronal functioning. These models tend to be uniform, focusing on the isolated penumbra, and typically lack adjacent regions such as a core and unaffected regions (normal perfusion). However, processes in these regions may affect neuronal functioning and survival in the penumbra. Materials and methods Here, we designed, fabricated, and characterized a cytocompatible device that generates an oxygen gradient across in vitro neuronal cultures to expose cells to hypoxia of various depths from near anoxia to near normoxia. This marks a step in the path to mimic core, penumbra, and healthy tissue, and will facilitate better in vitro modeling of ischemic stroke. Results The generator forms a stable and reproducible gradient within 30 min. Oxygen concentrations at the extremes are adjustable in a physiologically relevant range. Application of the generator did not negatively affect electrophysiological recordings or the viability of cultures, thus confirming the cytocompatibility of the device. Discussion The developed device is able to impose an oxygen gradient on neuronal cultures and may enrich in vitro stroke models.

Theory suggest that networks of neurons may predict their input. Prediction may underlie most aspects of information processing and is believed to be involved in motor and cognitive control and decision-making. Retinal cells have been shown to be capable of predicting visual stimuli, and there is some evidence for prediction of input in the visual cortex and hippocampus. However, there is no proof that the ability to predict is a generic feature of neural networks. We investigated whether random in vitro neuronal networks can predict stimulation, and how prediction is related to short- and long-term memory. To answer these questions, we applied two different stimulation modalities. Focal electrical stimulation has been shown to induce long-term memory traces, whereas global optogenetic stimulation did not. We used mutual information to quantify how much activity recorded from these networks reduces the uncertainty of upcoming stimuli (prediction) or recent past stimuli (short-term memory). Cortical neural networks did predict future stimuli, with the majority of all predictive information provided by the immediate network response to the stimulus. Interestingly, prediction strongly depended on short-term memory of recent sensory inputs during focal as well as global stimulation. However, prediction required less short-term memory during focal stimulation. Furthermore, the dependency on short-term memory decreased during 20 h of focal stimulation, when long-term connectivity changes were induced. These changes are fundamental for long-term memory formation, suggesting that besides short-term memory the formation of long-term memory traces may play a role in efficient prediction.

In the penumbra of a brain infarct, neurons initially remain structurally intact, but perfusion is insufficient to maintain neuronal activity at physiological levels. Improving neuronal recovery in the penumbra has large potential to advance recovery of stroke patients, but penumbral pathology is incompletely understood, and treatments are scarce. We hypothesize that low activity in the penumbra is associated with apoptosis and thus contributes to irreversible neuronal damage. We explored the putative relationship between low neuronal activity and apoptosis in cultured neurons exposed to variable durations of hypoxia or TTX. We combined electrophysiology and live apoptosis staining in 42 cultures, and compared effects of hypoxia and TTX silencing in terms of network activity and apoptosis. Hypoxia rapidly reduced network activity, but cultures showed limited apoptosis during the first 12 h. After 24 h, widespread apoptosis had occurred. This was associated with full activity recovery observed upon reoxygenation within 12 h, but not after 24 h. Similarly, TTX exposure strongly reduced activity, with full recovery upon washout within 12 h, but not after 24 h. Mean temporal evolution of apoptosis in TTX-treated cultures was the same as in hypoxic cultures. These results suggest that prolonged low activity may be a common factor in the pathways towards apoptosis.

Tools to estimate brain connectivity offer the potential to enhance our understanding of brain functioning. The behavior of neuronal networks, including functional connectivity and induced connectivity changes by external stimuli, can be studied using models of cultured neurons. Cultured neurons tend to be active in groups, and pairs of neurons are said to be functionally connected when their firing patterns show significant synchronicity. Methods to infer functional connections are often based on pair-wise cross-correlation between activity patterns of (small groups of) neurons. However, these methods are not very sensitive to detect inhibitory connections, and they were not designed for use during stimulation. Maximum Entropy (MaxEnt) models may provide a conceptually different method to infer functional connectivity. They have the potential benefit to estimate functional connectivity during stimulation, and to infer excitatory as well as inhibitory connections. MaxEnt models do not involve pairwise comparison, but aim to capture probability distributions of sets of neurons that are synchronously active in discrete time bins. We used electrophysiological recordings from in vitro neuronal cultures on micro electrode arrays to investigate the ability of MaxEnt models to infer functional connectivity. Connectivity estimates provided by MaxEnt models correlated well with those obtained by conditional firing probabilities (CFP), an established cross-correlation based method. In addition, stimulus-induced connectivity changes were detected by MaxEnt models, and were of the same magnitude as those detected by CFP. Thus, MaxEnt models provide a potentially powerful new tool to study functional connectivity in neuronal networks.

Functional assessment of in vitro neuronal networks—of relevance for disease modelling and drug testing—can be performed using multi-electrode array (MEA) technology. However, the handling and processing of the large amount of data typically generated in MEA experiments remains a huge hurdle for researchers. Various software packages have been developed to tackle this issue, but to date, most are either not accessible through the links provided by the authors or only tackle parts of the analysis. Here, we present ‘‘ MEA-ToolBox ’’, a free open-source general MEA analytical toolbox that uses a variety of literature-based algorithms to process the data, detect spikes from raw recordings, and extract information at both the single-channel and array-wide network level. MEA-ToolBox extracts information about spike trains, burst-related analysis and connectivity metrics without the need of manual intervention. MEA-ToolBox is tailored for comparing different sets of measurements and will analyze data from multiple recorded files placed in the same folder sequentially, thus considerably streamlining the analysis pipeline. MEA-ToolBox is available with a graphic user interface (GUI) thus eliminating the need for any coding expertise while offering functionality to inspect, explore and post-process the data. As proof-of-concept, MEA-ToolBox was tested on earlier-published MEA recordings from neuronal networks derived from human induced pluripotent stem cells (hiPSCs) obtained from healthy subjects and patients with neurodevelopmental disorders. Neuronal networks derived from patient’s hiPSCs showed a clear phenotype compared to those from healthy subjects, demonstrating that the toolbox could extract useful parameters and assess differences between normal and diseased profiles.

The cerebral pressure reactivity index (PRx), through intracranial pressure (ICP) measurements, informs clinicians about the cerebral autoregulation (CA) status in adult-sedated patients with traumatic brain injury (TBI). Using PRx in clinical practice is currently limited by variability over shorter monitoring periods. We applied an innovative method to reduce the PRx variability by ventilator-induced slow (1/min) positive end-expiratory pressure (PEEP) oscillations. We hypothesized that, as seen in a previous animal model, the PRx variability would be reduced by inducing slow arterial blood pressure (ABP) and ICP oscillations without other clinically relevant physiological changes. Patients with TBI were ventilated with a static PEEP for 30 min (PRx period) followed by a 30-min period of slow [1/min (0.0167 Hz)] +5 cmH2O PEEP oscillations (induced (iPRx period). Ten patients with TBI were included. No clinical monitoring was discontinued and no additional interventions were required during the iPRx period. The PRx variability [measured as the standard deviation (SD) of PRx] decreased significantly during the iPRx period from 0.25 (0.22-0.30) to 0.14 (0.09-0.17) (P = 0.006). There was a power increase around the induced frequency (1/min) for both ABP and ICP (P = 0.002). In conclusion, 1/min PEEP-induced oscillations reduced the PRx variability in patients with TBI with ICP levels <22 mmHg. No other clinically relevant physiological changes were observed. Reduced PRx variability might improve CA-guided perfusion management by reducing the time to find "optimal" perfusion pressure targets. Larger studies with prolonged periods of PEEP-induced oscillations are required to take it to routine use.NEW & NOTEWORTHY Cerebral autoregulation assessment requires sufficient slow arterial blood pressure (ABP) waves. However, spontaneous ABP waves may be insufficient for reliable cerebral autoregulation estimations. Therefore, we applied a ventilator "sigh-function" to generate positive end-expiratory pressure oscillations that induce slow ABP waves. This method demonstrated a reduced variability of the pressure reactivity index, commonly used as continuous cerebral autoregulation measure in a traumatic brain injury population.

Postanoxic encephalopathy is the key determinant of death or disability after successful cardiopulmonary resuscitation. Animal studies have provided proof-of-principle evidence of efficacy of divergent classes of neuroprotective treatments to promote brain recovery. However, apart from targeted temperature management (TTM), neuroprotective treatments are not included in current care of patients with postanoxic encephalopathy after cardiac arrest. We aimed to review the clinical evidence of efficacy of neuroprotective strategies to improve recovery of comatose patients after cardiac arrest and to propose future directions. We performed a systematic search of the literature to identify prospective, comparative clinical trials on interventions to improve neurological outcome of comatose patients after cardiac arrest. We included 53 studies on 21 interventions. None showed unequivocal benefit. TTM at 33 or 36°C and adrenaline (epinephrine) are studied most, followed by xenon, erythropoietin, and calcium antagonists. Lack of efficacy is associated with heterogeneity of patient groups and limited specificity of outcome measures. Ongoing and future trials will benefit from systematic collection of measures of baseline encephalopathy and sufficiently powered predefined subgroup analyses. Outcome measurement should include comprehensive neuropsychological follow-up, to show treatment effects that are not detectable by gross measures of functional recovery. To enhance translation from animal models to patients, studies under experimental conditions should adhere to strict methodological and publication guidelines.

OBJECTIVE:In ischemic stroke, treatments to protect neurons from irreversible damage are urgently needed. Studies in animal models have shown that neuroprotective treatments targeting neuronal silencing improve brain recovery, but in clinical trials none of these were effective in patients. This failure of translation poses doubts on the real efficacy of treatments tested and on the validity of animal models for human stroke. Here, we established a human neuronal model of the ischemic penumbra by using human induced pluripotent stem cells and we provided an in-depth characterization of neuronal responses to hypoxia and treatment strategies at the network level.APPROACH:We generated neurons from induced pluripotent stem cells derived from healthy donor and we cultured them on micro-electrode arrays. We measured the electrophysiological activity of human neuronal networks under controlled hypoxic conditions. We tested the effect of different treatment strategies on neuronal network functionality.MAIN RESULTS:Human neuronal networks are vulnerable to hypoxia reflected by a decrease in activity and synchronicity under low oxygen conditions. We observe that full, partial or absent recovery depend on the timing of re-oxygenation and we provide a critical time threshold that, if crossed, is associated with irreversible impairments. We found that hypoxic preconditioning improves resistance to a second hypoxic insult. Finally, in contrast to previously tested, ineffective treatments, we show that stimulatory treatments counteracting neuronal silencing during hypoxia, such as optogenetic stimulation, are neuroprotective.SIGNIFICANCE:We presented a human neuronal model of the ischemic penumbra and we provided insights that may offer the basis for novel therapeutic approaches for patients after stroke. The use of human neurons might improve drug discovery and translation of findings to patients and might open new perspectives for personalized investigations.

In systems consolidation, encoded memories are replayed by the hippocampus during slow-wave sleep (SWS), and permanently stored in the neocortex. Declarative memory consolidation is believed to benefit from the oscillatory rhythms and low cholinergic tone observed in this sleep stage, but underlying mechanisms remain unclear. To clarify the role of cholinergic modulation and synchronized activity in memory consolidation, we applied repeated electrical stimulation in mature cultures of dissociated rat cortical neurons with high or low cholinergic tone, mimicking the cue replay observed during systems consolidation under distinct cholinergic concentrations. In the absence of cholinergic input, these cultures display activity patterns hallmarked by network bursts, synchronized events reminiscent of the low frequency oscillations observed during SWS. They display stable activity and connectivity, which mutually interact and achieve an equilibrium. Electrical stimulation reforms the equilibrium to include the stimulus response, a phenomenon interpreted as memory trace formation. Without cholinergic input, activity was burst-dominated. First application of a stimulus induced significant connectivity changes, while subsequent repetition no longer affected connectivity. Presenting a second stimulus at a different electrode had the same effect, whereas returning to the initial stimuli did not induce further connectivity alterations, indicating that the second stimulus did not erase the 'memory trace' of the first. Distinctively, cultures with high cholinergic tone displayed reduced network excitability and dispersed firing, and electrical stimulation did not induce significant connectivity changes. We conclude that low cholinergic tone facilitates memory formation and consolidation, possibly through enhanced network excitability. Network bursts or SWS oscillations may merely reflect high network excitability.

The brain is the most complex organ of the body, and many pathological processes underlying various brain disorders are poorly understood. Limited accessibility hinders observation of such processes in the in vivo brain, and experimental freedom is often insufficient to enable informative manipulations. In vitro preparations (brain slices or cultures of dissociated neurons) offer much better accessibility and reduced complexity and have yielded valuable new insights into various brain disorders. Both types of preparations have their advantages and limitations with regard to lifespan, preservation of in vivo brain structure, composition of cell types, and the link to behavioral outcome is often unclear in in vitro models. While these limitations hamper general usage of in vitro preparations to study, e.g., brain development, in vitro preparations are very useful to study neuronal and synaptic functioning under pathologic conditions. This chapter addresses several brain disorders, focusing on neuronal and synaptic functioning, as well as network aspects. Recent progress in the fields of brain circulation disorders, excitability disorders, and memory disorders will be discussed, as well as limitations of current in vitro models.