Performance investigation of epilepsy detection from noisy EEG signals using base-2-meta stacking classifier.

Epilepsy is a chronic neurological disease, characterized by spontaneous, unprovoked, recurrent seizures that may lead to long-term disability and premature death. Despite significant efforts made to improve epilepsy detection clinically and pre-clinically, the pervasive presence of noise in EEG signals continues to pose substantial challenges to their effective application. In addition, discriminant features for epilepsy detection have not been investigated yet. The objective of this study is to develop a hybrid model for epilepsy detection from noisy and fragmented EEG signals. We hypothesized that a hybrid model could surpass existing single models in epilepsy detection. Our approach involves manual noise rejection and a novel statistical channel selection technique to detect epilepsy even from noisy EEG signals. Our proposed Base-2-Meta stacking classifier achieved notable accuracy (0.98 ± 0.05), precision (0.98 ± 0.07), recall (0.98 ± 0.05), and F1 score (0.98 ± 0.04) even with noisy 5-s segmented EEG signals. Application of our approach to the specific problem like detection of epilepsy from noisy and fragmented EEG data reveals a performance that is not only superior to others, but also is translationally relevant, highlighting its potential application in a clinic setting, where EEG signals are often noisy or scanty. Our proposed metric DF-A (Discriminant feature-accuracy), for the first time, identified the most discriminant feature with models that give A accuracy or above (A = 95 used in this study). This groundbreaking approach allows for detecting discriminant features and can be used as potential electrographic biomarkers in epilepsy detection research. Moreover, our study introduces innovative insights into the understanding of these features, epilepsy detection, and cross-validation, markedly improving epilepsy detection in ways previously unavailable.

EEG power spectra parameterization and adaptive channel selection towards semi-supervised seizure prediction.

BACKGROUND: The seizure prediction algorithms have demonstrated their potential in mitigating epilepsy risks by detecting the pre-ictal state using ongoing electroencephalogram (EEG) signals. However, most of them require high-density EEG, which is burdensome to the patients for daily monitoring. Moreover, prevailing seizure models require extensive training with significant labeled data which is very time-consuming and demanding for the epileptologists. METHOD: To address these challenges, here we propose an adaptive channel selection strategy and a semi-supervised deep learning model respectively to reduce the number of EEG channels and to limit the amount of labeled data required for accurate seizure prediction. Our channel selection module is centered on features from EEG power spectra parameterization that precisely characterize the epileptic activities to identify the seizure-associated channels for each patient. The semi-supervised model integrates generative adversarial networks and bidirectional long short-term memory networks to enhance seizure prediction. RESULTS: Our approach is evaluated on the CHB-MIT and Siena epilepsy datasets. With utilizing only 4 channels, the method demonstrates outstanding performance with an AUC of 93.15% on the CHB-MIT dataset and an AUC of 88.98% on the Siena dataset. Experimental results also demonstrate that our selection approach reduces the model parameters and training time. CONCLUSIONS: Adaptive channel selection coupled with semi-supervised learning can offer the possible bases for a light weight and computationally efficient seizure prediction system, making the daily monitoring practical to improve patients' quality of life.

Feasibility of decoding covert speech in ECoG with a Transformer trained on overt speech.

Several attempts for speech brain-computer interfacing (BCI) have been made to decode phonemes, sub-words, words, or sentences using invasive measurements, such as the electrocorticogram (ECoG), during auditory speech perception, overt speech, or imagined (covert) speech. Decoding sentences from covert speech is a challenging task. Sixteen epilepsy patients with intracranially implanted electrodes participated in this study, and ECoGs were recorded during overt speech and covert speech of eight Japanese sentences, each consisting of three tokens. In particular, Transformer neural network model was applied to decode text sentences from covert speech, which was trained using ECoGs obtained during overt speech. We first examined the proposed Transformer model using the same task for training and testing, and then evaluated the model's performance when trained with overt task for decoding covert speech. The Transformer model trained on covert speech achieved an average token error rate (TER) of 46.6% for decoding covert speech, whereas the model trained on overt speech achieved a TER of 46.3% ( p > 0.05 ; d = 0.07 ) . Therefore, the challenge of collecting training data for covert speech can be addressed using overt speech. The performance of covert speech can improve by employing several overt speeches.

Differentiating Epileptic and Psychogenic Non-Epileptic Seizures Using Machine Learning Analysis of EEG Plot Images.

The treatment of epilepsy, the second most common chronic neurological disorder, is often complicated by the failure of patients to respond to medication. Treatment failure with anti-seizure medications is often due to the presence of non-epileptic seizures. Distinguishing non-epileptic from epileptic seizures requires an expensive and time-consuming analysis of electroencephalograms (EEGs) recorded in an epilepsy monitoring unit. Machine learning algorithms have been used to detect seizures from EEG, typically using EEG waveform analysis. We employed an alternative approach, using a convolutional neural network (CNN) with transfer learning using MobileNetV2 to emulate the real-world visual analysis of EEG images by epileptologists. A total of 5359 EEG waveform plot images from 107 adult subjects across two epilepsy monitoring units in separate medical facilities were divided into epileptic and non-epileptic groups for training and cross-validation of the CNN. The model achieved an accuracy of 86.9% (Area Under the Curve, AUC 0.92) at the site where training data were extracted and an accuracy of 87.3% (AUC 0.94) at the other site whose data were only used for validation. This investigation demonstrates the high accuracy achievable with CNN analysis of EEG plot images and the robustness of this approach across EEG visualization software, laying the groundwork for further subclassification of seizures using similar approaches in a clinical setting.

Calibrating Deep Learning Classifiers for Patient-Independent Electroencephalogram Seizure Forecasting.

The recent scientific literature abounds in proposals of seizure forecasting methods that exploit machine learning to automatically analyze electroencephalogram (EEG) signals. Deep learning algorithms seem to achieve a particularly remarkable performance, suggesting that the implementation of clinical devices for seizure prediction might be within reach. However, most of the research evaluated the robustness of automatic forecasting methods through randomized cross-validation techniques, while clinical applications require much more stringent validation based on patient-independent testing. In this study, we show that automatic seizure forecasting can be performed, to some extent, even on independent patients who have never been seen during the training phase, thanks to the implementation of a simple calibration pipeline that can fine-tune deep learning models, even on a single epileptic event recorded from a new patient. We evaluate our calibration procedure using two datasets containing EEG signals recorded from a large cohort of epileptic subjects, demonstrating that the forecast accuracy of deep learning methods can increase on average by more than 20%, and that performance improves systematically in all independent patients. We further show that our calibration procedure works best for deep learning models, but can also be successfully applied to machine learning algorithms based on engineered signal features. Although our method still requires at least one epileptic event per patient to calibrate the forecasting model, we conclude that focusing on realistic validation methods allows to more reliably compare different machine learning approaches for seizure prediction, enabling the implementation of robust and effective forecasting systems that can be used in daily healthcare practice.

Diagnostic biomarker discovery from brain EEG data using LSTM, reservoir-SNN, and NeuCube methods in a pilot study comparing epilepsy and migraine.

The study introduces a new online spike encoding algorithm for spiking neural networks (SNN) and suggests new methods for learning and identifying diagnostic biomarkers using three prominent deep learning neural network models: deep BiLSTM, reservoir SNN, and NeuCube. EEG data from datasets related to epilepsy, migraine, and healthy subjects are employed. Results reveal that BiLSTM hidden neurons capture biological significance, while reservoir SNN activities and NeuCube spiking dynamics identify EEG channels as diagnostic biomarkers. BiLSTM and reservoir SNN achieve 90 and 85% classification accuracy, while NeuCube achieves 97%, all methods pinpointing potential biomarkers like T6, F7, C4, and F8. The research bears implications for refining online EEG classification, analysis, and early brain state diagnosis, enhancing AI models with interpretability and discovery. The proposed techniques hold promise for streamlined brain-computer interfaces and clinical applications, representing a significant advancement in pattern discovery across the three most popular neural network methods for addressing a crucial problem. Further research is planned to study how early can these diagnostic biomarkers predict an onset of brain states.

Seizure Outcome After Intraoperative Electrocorticography-Tailored Epilepsy Surgery: A Systematic Review and Meta-Analysis.

BACKGROUND AND OBJECTIVES: Tailoring epilepsy surgery using intraoperative electrocorticography (ioECoG) has been debated, and modest number of epilepsy surgery centers apply this diagnostic method. We assessed the current evidence to use ioECoG-tailored epilepsy surgery for improving postsurgical outcome. METHODS: PubMed and Embase were searched for original studies reporting on ≥10 cases who underwent ioECoG-tailored surgery for epilepsy, with a follow-up of at least 6 months. We used a random-effects model to calculate the overall rate of patients achieving favorable seizure outcome (FSO), defined as Engel class I, ILAE class 1, or seizure-free status. Meta-regression was used to investigate potential sources of heterogeneity. We calculated the odds ratio (OR) for estimating variables on FSO:ioECoG vs non-ioECoG-tailored surgery (if included studies contained patients with non-ioECoG-tailored surgery), ioECoG-tailored epilepsy surgery in children vs adults, temporal (TL) vs extratemporal lobe (eTL), MRI-positive vs MRI-negative, and complete vs incomplete resection of tissue that generated interictal epileptiform discharges (IEDs). A Bayesian network meta-analysis was conducted for underlying pathologies. We assessed the evidence certainty using the Grading of Recommendations, Assessment, Development, and Evaluation (GRADE). RESULTS: Eighty-three studies (82 observational studies, 1 trial) comprising 3,631 patients with ioECoG-tailored surgery were included. The overall pooled rate of patients who attained FSO after ioECoG-tailored surgery was 74% (95% CI 71-77) with significant heterogeneity, which was predominantly attributed to pathologies and seizure outcome classifications. Twenty-two studies contained non-ioECoG-tailored surgeries. IoECoG-tailored surgeries reached a higher rate of FSO than non-ioECoG-tailored surgeries (OR 2.10 [95% CI 1.37-3.24]; p < 0.01; very low certainty). Complete resection of tissue that displayed IEDs in ioECoG predicted FSO better compared with incomplete resection (OR 3.04 [1.76-5.25]; p < 0.01; low certainty). We found insignificant difference in FSO after ioECoG-tailored surgery in children vs adults, TL vs eTL, or MRI-positive vs MRI-negative. The network meta-analysis showed that the odds of FSO was lower for malformations of cortical development than for tumors (OR 0.47 95% credible interval 0.25-0.87). DISCUSSION: Although limited by low-quality evidence, our meta-analysis shows a relatively good surgical outcome (74% FSO) after epilepsy surgery with ioECoG, especially in tumors, with better outcome for ioECoG-tailored surgeries in studies describing both and better outcome after complete removal of IED areas.

Prevalence of abnormal electrocardiographic findings and associated factors among patients with epilepsy at Jimma Medical Center, Ethiopia; Hospital-based cross-sectional study.

BACKGROUND: A person with epilepsy experiences recurrent seizures as a result of a persistent underlying disorder. About 50 million people globally are impacted by it, with 4 million of those being in Sub-Saharan Africa. One of the most frequent comorbidities that raise the mortality and morbidity rates of epileptic patients is abnormal Electrocardiographic (ECG) findings. Thus, the purpose of this study is to evaluate the prevalence of abnormal ECG findings in epileptic patients that might lead to increased risk of sudden cardiac death. METHODOLOGY: A hospital based cross-sectional study was at Jimma Medical Center of Ethiopia on epileptic patients who were on follow-up at neurologic clinics during the data collection period. The malignant ECG characteristics and was identified using the ECG abnormality tool. To facilitate analysis, the gathered data was imported into Epidata version 3.1 and exported to the SPSS version 26. The factors of abnormal ECG and sudden death risk were examined using bivariate logistic regression. RESULTS: The study comprised 190 epileptic patients, with a mean age of 32 years. There were more men than women, making up 60.2%. A 43.2% (n = 80) frequency of ECG abnormalities was identified. According to the study, early repolarization abnormalities were the most common ECG abnormalities and increased with male sex and the length of time a person had seizures (AOR) of 4.751 and 95% CI (.273,.933), p = 0.029, compared to their female counterparts. CONCLUSION: The frequency of malignant ECG alterations in epileptic patients on follow-up at Jimma Medical Center in Ethiopia is described in the study. According to the study, there were significant ECG alterations in epileptic individuals. Male gender and longer duration of epilepsy raise the risk of abnormal ECG findings that could result in sudden cardiac death.

Aperiodic components of local field potentials reflect inherent differences between cortical and subcortical activity.

Information flow in brain networks is reflected in local field potentials that have both periodic and aperiodic components. The 1/fχ aperiodic component of the power spectra tracks arousal and correlates with other physiological and pathophysiological states. Here we explored the aperiodic activity in the human thalamus and basal ganglia in relation to simultaneously recorded cortical activity. We elaborated on the parameterization of the aperiodic component implemented by specparam (formerly known as FOOOF) to avoid parameter unidentifiability and to obtain independent and more easily interpretable parameters. This allowed us to seamlessly fit spectra with and without an aperiodic knee, a parameter that captures a change in the slope of the aperiodic component. We found that the cortical aperiodic exponent χ, which reflects the decay of the aperiodic component with frequency, is correlated with Parkinson's disease symptom severity. Interestingly, no aperiodic knee was detected from the thalamus, the pallidum, or the subthalamic nucleus, which exhibited an aperiodic exponent significantly lower than in cortex. These differences were replicated in epilepsy patients undergoing intracranial monitoring that included thalamic recordings. The consistently lower aperiodic exponent and lack of an aperiodic knee from all subcortical recordings may reflect cytoarchitectonic and/or functional differences. SIGNIFICANCE STATEMENT: The aperiodic component of local field potentials can be modeled to produce useful and reproducible indices of neural activity. Here we refined a widely used phenomenological model for extracting aperiodic parameters (namely the exponent, offset and knee), with which we fit cortical, basal ganglia, and thalamic intracranial local field potentials, recorded from unique cohorts of movement disorders and epilepsy patients. We found that the aperiodic exponent in motor cortex is higher in Parkinson's disease patients with more severe motor symptoms, suggesting that aperiodic features may have potential as electrophysiological biomarkers for movement disorders symptoms. Remarkably, we found conspicuous differences in the aperiodic parameters of basal ganglia and thalamic signals compared to those from neocortex.

Pathophysiology, Diagnosis, Prognosis, and Prevention of Poststroke Epilepsy: Clinical and Research Implications.

Poststroke epilepsy (PSE) is associated with higher mortality and poor functional and cognitive outcomes in patients with stroke. With the remarkable development of acute stroke treatment, there is a growing number of survivors with PSE. Although approximately 10% of patients with stroke develop PSE, given the significant burden of stroke worldwide, PSE is a significant problem in stroke survivors. Therefore, the attention of health policymakers and significant funding are required to promote PSE prevention research. The current PSE definition includes unprovoked seizures occurring more than 7 days after stroke onset, given the high recurrence risks of seizures. However, the pathologic cascade of stroke is not uniform, indicating the need for a tissue-based approach rather than a time-based one to distinguish early seizures from late seizures. EEG is a commonly used tool in the diagnostic work-up of PSE. EEG findings during the acute phase of stroke can potentially stratify the risk of subsequent seizures and predict the development of poststroke epileptogenesis. Recent reports suggest that cortical superficial siderosis, which may be involved in epileptogenesis, is a promising marker for PSE. By incorporating such markers, future risk-scoring models could guide treatment strategies, particularly for the primary prophylaxis of PSE. To date, drugs that prevent poststroke epileptogenesis are lacking. The primary challenge involves the substantial cost burden due to the difficulty of reliably enrolling patients who develop PSE. There is, therefore, a critical need to determine reliable biomarkers for PSE. The goal is to be able to use them for trial enrichment and as a surrogate outcome measure for epileptogenesis. Moreover, seizure prophylaxis is essential to prevent functional and cognitive decline in stroke survivors. Further elucidation of factors that contribute to poststroke epileptogenesis is eagerly awaited. Meanwhile, the regimen of antiseizure medications should be based on individual cardiovascular risk, psychosomatic comorbidities, and concomitant medications. This review summarizes the current understanding of poststroke epileptogenesis, its risks, prognostic models, prophylaxis, and strategies for secondary prevention of seizures and suggests strategies to advance research on PSE.

Contribution of microglia to the epileptiform activity that results from neonatal hypoxia.

Microglia are described as the immune cells of the brain, their immune properties have been extensively studied since first described, however, their neural functions have only been explored over the last decade. Microglia have an important role in maintaining homeostasis in the central nervous system by surveying their surroundings to detect pathogens or damage cells. While these are the classical functions described for microglia, more recently their neural functions have been defined; they are critical to the maturation of neurons during embryonic and postnatal development, phagocytic microglia remove excess synapses during development, a process called synaptic pruning, which is important to overall neural maturation. Furthermore, microglia can respond to neuronal activity and, together with astrocytes, can regulate neural activity, contributing to the equilibrium between excitation and inhibition through a feedback loop. Hypoxia at birth is a serious neurological condition that disrupts normal brain function resulting in seizures and epilepsy later in life. Evidence has shown that microglia may contribute to this hyperexcitability after neonatal hypoxia. This review will summarize the existing data on the role of microglia in the pathogenesis of neonatal hypoxia and the plausible mechanisms that contribute to the development of hyperexcitability after hypoxia in neonates. This article is part of the Special Issue on "Microglia".

A multiscale symbolic approach to decoding delta and ripple oscillation bands as biomarkers for epileptiform discharges.

We use a multiscale symbolic approach to study the complex dynamics of temporal lobe refractory epilepsy employing high-resolution intracranial electroencephalogram (iEEG). We consider the basal and preictal phases and meticulously analyze the dynamics across frequency bands, focusing on high-frequency oscillations up to 240 Hz. Our results reveal significant periodicities and critical time scales within neural dynamics across frequency bands. By bandpass filtering neural signals into delta, theta, alpha, beta, gamma, and ripple high-frequency bands (HFO), each associated with specific neural processes, we examine the distinct nonlinear dynamics. Our method introduces a reliable approach to pinpoint intrinsic time lag scales τ within frequency bands of the basal and preictal signals, which are crucial for the study of refractory epilepsy. Using metrics such as permutation entropy (H), Fisher information (F), and complexity (C), we explore nonlinear patterns within iEEG signals. We reveal the intrinsic τmax that maximize complexity within each frequency band, unveiling the nonlinear subtle patterns of the temporal structures within the basal and preictal signal. Examining the H×F and C×F values allows us to identify differences in the delta band and a band between 200 and 220 Hz (HFO 6) when comparing basal and preictal signals. Differences in Fisher information in the delta and HFO 6 bands before seizures highlight their role in capturing important system dynamics. This offers new perspectives on the intricate relationship between delta oscillations and HFO waves in patients with focal epilepsy, highlighting the importance of these patterns and their potential as biomarkers.

The value of linear and non-linear quantitative EEG analysis in paediatric epilepsy surgery: a machine learning approach.

Epilepsy surgery is effective for patients with medication-resistant seizures, however 20-40% of them are not seizure free after surgery. Aim of this study is to evaluate the role of linear and non-linear EEG features to predict post-surgical outcome. We included 123 paediatric patients who underwent epilepsy surgery at Bambino Gesù Children Hospital (January 2009-April 2020). All patients had long term video-EEG monitoring. We analysed 1-min scalp interictal EEG (wakefulness and sleep) and extracted 13 linear and non-linear EEG features (power spectral density (PSD), Hjorth, approximate entropy, permutation entropy, Lyapunov and Hurst value). We used a logistic regression (LR) as feature selection process. To quantify the correlation between EEG features and surgical outcome we used an artificial neural network (ANN) model with 18 architectures. LR revealed a significant correlation between PSD of alpha band (sleep), Mobility index (sleep) and the Hurst value (sleep and awake) with outcome. The fifty-four ANN models gave a range of accuracy (46-65%) in predicting outcome. Within the fifty-four ANN models, we found a higher accuracy (64.8% ± 7.6%) in seizure outcome prediction, using features selected by LR. The combination of PSD of alpha band, mobility and the Hurst value positively correlate with good surgical outcome.

vEpiNet: A multimodal interictal epileptiform discharge detection method based on video and electroencephalogram data.

To enhance deep learning-based automated interictal epileptiform discharge (IED) detection, this study proposes a multimodal method, vEpiNet, that leverages video and electroencephalogram (EEG) data. Datasets comprise 24 931 IED (from 484 patients) and 166 094 non-IED 4-second video-EEG segments. The video data is processed by the proposed patient detection method, with frame difference and Simple Keypoints (SKPS) capturing patients' movements. EEG data is processed with EfficientNetV2. The video and EEG features are fused via a multilayer perceptron. We developed a comparative model, termed nEpiNet, to test the effectiveness of the video feature in vEpiNet. The 10-fold cross-validation was used for testing. The 10-fold cross-validation showed high areas under the receiver operating characteristic curve (AUROC) in both models, with a slightly superior AUROC (0.9902) in vEpiNet compared to nEpiNet (0.9878). Moreover, to test the model performance in real-world scenarios, we set a prospective test dataset, containing 215 h of raw video-EEG data from 50 patients. The result shows that the vEpiNet achieves an area under the precision-recall curve (AUPRC) of 0.8623, surpassing nEpiNet's 0.8316. Incorporating video data raises precision from 70% (95% CI, 69.8%-70.2%) to 76.6% (95% CI, 74.9%-78.2%) at 80% sensitivity and reduces false positives by nearly a third, with vEpiNet processing one-hour video-EEG data in 5.7 min on average. Our findings indicate that video data can significantly improve the performance and precision of IED detection, especially in prospective real clinic testing. It suggests that vEpiNet is a clinically viable and effective tool for IED analysis in real-world applications.

Spatial and temporal properties of intra-operatively recorded spikes and high frequency oscillations in focal cortical dysplasia.

OBJECTIVE: Focal cortical dysplasias (FCD) are characterized by distinct interictal spike patterns and high frequency oscillations (HFOs; ripples: 80-250 Hz; fast ripples: 250-500 Hz) in the intra-operative electrocorticogram (ioECoG). We studied the temporal relation between intra-operative spikes and HFOs and their relation to resected tissue in people with FCD with a favorable outcome. METHODS: We included patients who underwent ioECoG-tailored epilepsy surgery with pathology confirmed FCD and long-term Engel 1A outcome. Spikes and HFOs were automatically detected and visually checked in 1-minute pre-resection-ioECoG. Channels covering resected and non-resected tissue were compared using a logistic mixed model, assessing event numbers, co-occurrence ratios, and time-based properties. RESULTS: We found pre-resection spikes, ripples in respectively 21 and 20 out of 22 patients. Channels covering resected tissue showed high numbers of spikes and HFOs, and high ratios of co-occurring events. Spikes, especially with ripples, have a relatively sharp rising flank with a long descending flank and early ripple onset over resected tissue. CONCLUSIONS: A combined analysis of event numbers, ratios, and temporal relationships between spikes and HFOs may aid identifying epileptic tissue in epilepsy surgery. SIGNIFICANCE: This study shows a promising method for clinically relevant properties of events, closely associated with FCD.

Deciphering the Enigma of Neuron-Glial Interactions in Neurological Disorders.

Innate lymphocytes, including microglial cells, astrocytes, and oligodendrocytes, play a crucial role in initiating neuroinflammatory reactions inside the central nervous system (CNS). The prime focus of this paper is on the involvement and interplay of neurons and glial cells in neurological disorders such as Alzheimer's Disease (AD), Autism Spectrum Disorder (ASD), epilepsy, and multiple sclerosis (MS). In this review, we explore the specific contributions of microglia and astrocytes and analyzes multiple pathways implicated in neuroinflammation and disturbances in excitatory and inhibitory processes. Firstly, we elucidate the mechanisms through which toxic protein accumulation in AD results in synaptic dysfunction and deregulation of the immune system and examines the roles of microglia, astrocytes, and hereditary factors in the pathogenesis of the disease. Secondly, we focus on ASD and the involvement of glial cells in the development of the nervous system and the formation of connections between neurons and investigates the genetic connections associated with these processes. Lastly, we also address the participation of glial cells in epilepsy and MS, providing insights into their pivotal functions in both conditions. We also tried to give an overview of seven different pathways like toll-like receptor signalling pathway, MyD88-dependent and independent pathway, etc and its relevance in the context with these neurological disorders. In this review, we also explore the role of activated glial cells in AD, ASD, epilepsy, and MS which lead to neuroinflammation. Even we focus on excitatory and inhibitory imbalance in all four neurological disorders as imbalance affect the proper functioning of neuronal circuits. Finally, this review concludes that there is necessity for additional investigation on glial cells and their involvement in neurological illnesses.

Control of neuronal excitation-inhibition balance by BMP-SMAD1 signalling.

Throughout life, neuronal networks in the mammalian neocortex maintain a balance of excitation and inhibition, which is essential for neuronal computation1,2. Deviations from a balanced state have been linked to neurodevelopmental disorders, and severe disruptions result in epilepsy3-5. To maintain balance, neuronal microcircuits composed of excitatory and inhibitory neurons sense alterations in neural activity and adjust neuronal connectivity and function. Here we identify a signalling pathway in the adult mouse neocortex that is activated in response to increased neuronal network activity. Overactivation of excitatory neurons is signalled to the network through an increase in the levels of BMP2, a growth factor that is well known for its role as a morphogen in embryonic development. BMP2 acts on parvalbumin-expressing (PV) interneurons through the transcription factor SMAD1, which controls an array of glutamatergic synapse proteins and components of perineuronal nets. PV-interneuron-specific disruption of BMP2-SMAD1 signalling is accompanied by a loss of glutamatergic innervation in PV cells, underdeveloped perineuronal nets and decreased excitability. Ultimately, this impairment of the functional recruitment of PV interneurons disrupts the cortical excitation-inhibition balance, with mice exhibiting spontaneous epileptic seizures. Our findings suggest that developmental morphogen signalling is repurposed to stabilize cortical networks in the adult mammalian brain.