Wavelet-based Gaussian-mixture hidden Markov model for the detection of multistage seizure dynamics: a proof-of-concept study.

BACKGROUND: Epilepsy is a common neurological disorder characterized by recurrent electrophysiological activities, known as seizures. Without the appropriate detection strategies, these seizure episodes can dramatically affect the quality of life for those afflicted. The rationale of this study is to develop an unsupervised algorithm for the detection of seizure states so that it may be implemented along with potential intervention strategies. METHODS: Hidden Markov model (HMM) was developed to interpret the state transitions of the in vitro rat hippocampal slice local field potentials (LFPs) during seizure episodes. It can be used to estimate the probability of state transitions and the corresponding characteristics of each state. Wavelet features were clustered and used to differentiate the electrophysiological characteristics at each corresponding HMM states. Using unsupervised training method, the HMM and the clustering parameters were obtained simultaneously. The HMM states were then assigned to the electrophysiological data using expert guided technique. Minimum redundancy maximum relevance (mRMR) analysis and Akaike Information Criterion (AICc) were applied to reduce the effect of over-fitting. The sensitivity, specificity and optimality index of chronic seizure detection were compared for various HMM topologies. The ability of distinguishing early and late tonic firing patterns prior to chronic seizures were also evaluated. RESULTS: Significant improvement in state detection performance was achieved when additional wavelet coefficient rates of change information were used as features. The final HMM topology obtained using mRMR and AICc was able to detect non-ictal (interictal), early and late tonic firing, chronic seizures and postictal activities. A mean sensitivity of 95.7%, mean specificity of 98.9% and optimality index of 0.995 in the detection of chronic seizures was achieved. The detection of early and late tonic firing was validated with experimental intracellular electrical recordings of seizures. CONCLUSIONS: The HMM implementation of a seizure dynamics detector is an improvement over existing approaches using visual detection and complexity measures. The subjectivity involved in partitioning the observed data prior to training can be eliminated. It can also decipher the probabilities of seizure state transitions using the magnitude and rate of change wavelet information of the LFPs.

Mechanisms of morphine enhancement of spontaneous seizure activity.

BACKGROUND: High-dose opioid therapy can precipitate seizures; however, the mechanism of such a dangerous adverse effect remains poorly understood. The aim of our study was to determine whether the neuroexcitatory activity of high-dose morphine is mediated by selective stimulation of opioid receptors. METHODS: Mice hippocampi were resected intact and bathed in low magnesium artificial cerebrospinal fluid to induce spontaneous seizure-like events recorded from CA1 neurons. RESULTS: Application of morphine had a biphasic effect on the recorded spontaneous seizure-like events. In a low concentration (10 microM), morphine depressed electrographic seizure activity. Higher morphine concentrations (30 and 100 microM) enhanced seizure activity in an apparent dose-dependent manner. Naloxone, a nonselective opiate antagonist blocked the proconvulsant action of morphine. Selective mu and kappa opiate receptor agonists and antagonists enhanced and suppressed the spontaneous seizure activity, respectively. On the contrary, delta opioid receptor ligands did not have an effect. CONCLUSIONS: The proseizure effect of morphine is mediated through selective stimulation of mu and kappa opiate receptors but not the activation of the delta receptor system. The observed dose-dependent mechanism of morphine neuroexcitation underscores careful adjustment and individualized opioid dosing in the clinical setting.

Lag synchrony measures dynamical processes underlying progression of seizure states.

We investigate the dynamics of bursting behavior in an intact hippocampal preparation using causal entropy, an adaptive measure of lag synchrony. This analysis, together with a heuristic model of coupled bursting networks, separates experimentally observed bursting dynamics into two dynamical regimes, when bursting is driven by (1) the intranetwork dynamics of a single region, or (2) internetwork feedback between spatially disjoint neural populations. Our results suggest that the abrupt transition between these two states heralds the gradual desynchronization of bursting activity. These results illustrate how superficially homogeneous behavior across loosely coupled networks may harbor hidden, but robust, dynamical processes.

Dynamics of epileptic phenomena determined from statistics of ictal transitions.

In this paper, we investigate the dynamical scenarios of transitions between normal and paroxysmal state in epilepsy. We assume that some epileptic neural network are bistable i.e., they feature two operational states, ictal and interictal that co-exist. The transitions between these two states may occur according to a Poisson process, a random walk process or as a result of deterministic time-dependent mechanisms. We analyze data from animal models of absence epilepsy, human epilepsies and in vitro models. The distributions of durations of ictal and interictal epochs are fitted with a gamma distribution. On the basis of qualitative features of the fits, we identify the dynamical processes that may have generated the underlying data. The analysis showed that the following hold. 1) The dynamics of ictal epochs differ from those of interictal states. 2) Seizure initiation can be accounted for by a random walk process while seizure termination is often mediated by deterministic mechanisms. 3) In certain cases, the transitions between ictal and interictal states can be modeled by a Poisson process operating in a bistable network. These results imply that exact prediction of seizure occurrence is not possible but termination of an ictal state by appropriate counter stimulation might be feasible.

Online prediction of onsets of seizure-like events in hippocampal neural networks using wavelet artificial neural networks.

It has been previously shown that wavelet artificial neural networks (WANNs) are able to classify the different states of epileptiform activity and predict the onsets of seizure-like events (SLEs) by offline processing (Ann. Biomed. Eng. 33(6):798-810, 2005) of the electrical data from the in-vitro hippocampal slice model of recurrent spontaneous SLEs. The WANN design entailed the assumption that time-varying frequency information from the biological recordings can be used to estimate the times at which onsets of SLEs would most likely occur in the future. Progressions of different frequency components were captured by the artificial neural network (ANN) using selective frequency inputs from the initial wavelet transform of the biological data. The training of the WANN had been established using 184 SLE episodes in 34 slices from 21 rats offline. Nine of these rats also exhibited periods of interictal bursts (IBs). These IBs were included as part of the training to help distinguish the difference in dynamics of bursting activities between the preictal- and interictal type. In this paper, we present the results of an online processing using WANN on 23 in-vitro rat hippocampal slices from 9 rats having 93 spontaneous SLE episodes generated under low magnesium conditions. Over the test cases, three of the nine rats exhibited over 30 min of IB activities. We demonstrated that the WANN was able to classify the different states, namely, interictal, preictal, ictal, and IB activities with an accuracy of 86.6, 72.6, 84.5, and 69.1%, respectively. Prediction of state transitions into ictal events was achieved using regression of initial "normalized time-to-onset" estimates. The SLE onsets can be estimated up to 36.4 s ahead of their actual occurrences, with a mean error of 14.3 +/- 27.0 s. The prediction errors decreased progressively as the actual time-to-onset decreased and more initial "normalized time-to-onset" estimates were used for the regression procedure.

Towards real-time in-implant epileptic seizure prediction.

We present an architecture of an epileptic seizure prediction system suitable for an implantable implementation. The microsystem comprises a neural interface, a spectral analysis processor and an artificial neural network (ANN). The neural interface and the spectral analysis processor have been prototyped in a 0.35 microm CMOS technology with experimental results are presented. The wavelet-based artificial neural network predicts the onsets of seizure up to two minutes before their occurrence in an in-vitro epilepsy model using a mouse hippocampal brain slice with recurrent spontaneous seizures.

Prevention of Spontaneous Seizure-like Events in Both in-silico and in-vitro Epilepsy Models.

In this paper, we apply the small perturbation control strategy for the prevention of seizure-like events (SLEs) characterized as lower dimensional possibly rhythmic (LPR) activities in both the coupled oscillators in-silico model and the in-vitro low magnesium rat hippocampal slice model. Utilizing the wavelet artificial neural network (WANN), state transitions towards SLEs can be predicted. Successful suppression of SLEs was achieved when brief control perturbations were applied to the field coupling portals of the coupled oscillators model and to the mossy fibers via extracellular field stimulating electrode, respectively.