Forecasting generalized epileptic seizures from the EEG signal by wavelet analysis and dynamic unsupervised fuzzy clustering

SDA: a data-driven algorithm that detects functional states applied to the EEG of Guhyasamaja meditation

The study presents a novel approach designed to detect time-continuous states in time-series data, called the State-Detecting Algorithm (SDA). The SDA operates on unlabeled data and detects optimal change-points among intrinsic functional states in time-series data based on an ensemble of Ward's hierarchical clustering with time-connectivity constraint. The algorithm chooses the best number of states and optimal state boundaries, maximizing clustering quality metrics. We also introduce a series of methods to estimate the performance and confidence of the SDA when the ground truth annotation is unavailable. These include information value analysis, paired statistical tests, and predictive modeling analysis. The SDA was validated on EEG recordings of Guhyasamaja meditation practice with a strict staged protocol performed by three experienced Buddhist practitioners in an ecological setup. The SDA used neurophysiological descriptors as inputs, including PSD, power indices, coherence, and PLV. Post-hoc analysis of the obtained EEG states revealed significant differences compared to the baseline and neighboring states. The SDA was found to be stable with respect to state order organization and showed poor clustering quality metrics and no statistical significance between states when applied to randomly shuffled epochs (i.e., surrogate subject data used as controls). The SDA can be considered a general data-driven approach that detects hidden functional states associated with the mental processes evolving during meditation or other ongoing mental and cognitive processes.

Preictal neuronal and vascular activity precedes the onset of childhood absence seizure: direct current potential shifts and their correlation with hemodynamic activity

SIGNIFICANCE AIMS

The neurovascular mechanisms underlying the initiation of absence seizures and their dynamics are still not well understood. The objective of this study was to better noninvasively characterize the dynamics of the neuronal and vascular network at the transition from the interictal state to the ictal state of absence seizures and back to the interictal state using a combined electroencephalography (EEG), functional near-infrared spectroscopy (fNIRS), and diffuse correlation spectroscopy (DCS) approach. The second objective was to develop hypotheses about the neuronal and vascular mechanisms that propel the networks to the 3-Hz spikes and wave discharges (SWDs) observed during absence seizures.

APPROACHES

We evaluated the simultaneous changes in electrical (neuronal) and optical dynamics [hemodynamic, with changes in (Hb) and cerebral blood flow] of 8 pediatric patients experiencing 25 typical childhood absence seizures during the transition from the interictal state to the absence seizure by simultaneously performing EEG, fNIRS, and DCS.

RESULTS

Starting from ∼ 20    s before the onset of the SWD, we observed a transient direct current potential shift that correlated with alterations in functional fNIRS and DCS measurements of the cerebral hemodynamics detecting the preictal changes.

DISCUSSION

Our noninvasive multimodal approach highlights the dynamic interactions between the neuronal and vascular compartments that take place in the neuronal network near the time of the onset of absence seizures in a very specific cerebral hemodynamic environment. These noninvasive approaches contribute to a better understanding of the electrical hemodynamic environment prior to seizure onset. Whether this may ultimately be relevant for diagnostic and therapeutic approaches requires further evaluation.

G., Murthy T. V., Bukhari SAC. EEG Forecasting With Univariate and Multivariate Time Series Using Windowing and Baseline Method

People suffering from epilepsy disorder are very much in need for precautionary measures. The only way to provide precaution to such people is to find some methods which help them to know in advance the occurrence of seizures. Using Electroencephalogram, the authors have worked on developing a forecasting method using simple LSTM with windowing technique. The window length was set to five time steps; step by step the length was increased by 1 time step. The number of correct predictions increased with the window length. When the length reached to 20 time steps, the model gave impressive results in predicting the future EEG value. Past 20 time steps are learnt by the neural network to forecast the future EEG in two stages; in univariate method, only one attribute is used as the basis to predict the future value. In multivariate method, 42 features were used to predict the future EEG. Multivariate is more powerful and provides the prediction which is almost equal to the actual target value. In case of univariate the accuracy achieved was about 70%, whereas in case of multivariate method it was 90%.

Connectivity maps based analysis of EEG for the advanced diagnosis of schizophrenia attributes

This article presents a novel connectivity analysis method that is suitable for multi-node networks such as EEG, MEG or EcOG electrode recordings. Its diagnostic power and ability to interpret brain states in schizophrenia is demonstrated on a set of 50 subjects that constituted of 25 healthy and 25 diagnosed with schizophrenia and treated with medication. The method can also be used for the automatic detection of schizophrenia; it exhibits higher sensitivity than state-of-the-art methods with no false positives. The detection is based on an analysis from a minute long pattern-recognition computer task. Moreover, this connectivity analysis leads naturally to an optimal choice of electrodes and hence to highly statistically significant results that are based on data from only 3-5 electrodes. The method is general and can be used for the diagnosis of other psychiatric conditions, provided an appropriate computer task is devised.

Schizophrenia detection and classification by advanced analysis of EEG recordings using a single electrode approach

Electroencephalographic (EEG) analysis has emerged as a powerful tool for brain state interpretation and diagnosis, but not for the diagnosis of mental disorders; this may be explained by its low spatial resolution or depth sensitivity. This paper concerns the diagnosis of schizophrenia using EEG, which currently suffers from several cardinal problems: it heavily depends on assumptions, conditions and prior knowledge regarding the patient. Additionally, the diagnostic experiments take hours, and the accuracy of the analysis is low or unreliable. This article presents the "TFFO" (Time-Frequency transformation followed by Feature-Optimization), a novel approach for schizophrenia detection showing great success in classification accuracy with no false positives. The methodology is designed for single electrode recording, and it attempts to make the data acquisition process feasible and quick for most patients.

Cluster-based exposure variation analysis

Background

Static posture, repetitive movements and lack of physical variation are known risk factors for work-related musculoskeletal disorders, and thus needs to be properly assessed in occupational studies. The aims of this study were (i) to investigate the effectiveness of a conventional exposure variation analysis (EVA) in discriminating exposure time lines and (ii) to compare it with a new cluster-based method for analysis of exposure variation.

Methods

For this purpose, we simulated a repeated cyclic exposure varying within each cycle between “low” and “high” exposure levels in a “near” or “far” range, and with “low” or “high” velocities (exposure change rates). The duration of each cycle was also manipulated by selecting a “small” or “large” standard deviation of the cycle time. Theses parameters reflected three dimensions of exposure variation, i.e. range, frequency and temporal similarity.

Each simulation trace included two realizations of 100 concatenated cycles with either low (ρ = 0.1), medium (ρ = 0.5) or high (ρ = 0.9) correlation between the realizations. These traces were analyzed by conventional EVA, and a novel cluster-based EVA (C-EVA). Principal component analysis (PCA) was applied on the marginal distributions of 1) the EVA of each of the realizations (univariate approach), 2) a combination of the EVA of both realizations (multivariate approach) and 3) C-EVA. The least number of principal components describing more than 90% of variability in each case was selected and the projection of marginal distributions along the selected principal component was calculated. A linear classifier was then applied to these projections to discriminate between the simulated exposure patterns, and the accuracy of classified realizations was determined.

Results

C-EVA classified exposures more correctly than univariate and multivariate EVA approaches; classification accuracy was 49%, 47% and 52% for EVA (univariate and multivariate), and C-EVA, respectively (p < 0.001). All three methods performed poorly in discriminating exposure patterns differing with respect to the variability in cycle time duration.

Conclusion

While C-EVA had a higher accuracy than conventional EVA, both failed to detect differences in temporal similarity. The data-driven optimality of data reduction and the capability of handling multiple exposure time lines in a single analysis are the advantages of the C-EVA.

Seizure prediction using adaptive neuro-fuzzy inference system

In this study, we present a neuro-fuzzy approach of seizure prediction from invasive Electroencephalogram (EEG) by applying adaptive neuro-fuzzy inference system (ANFIS). Three nonlinear seizure predictive features were extracted from a patient's data obtained from the European Epilepsy Database, one of the most comprehensive EEG database for epilepsy research. A total of 36 hours of recordings including 7 seizures was used for analysis. The nonlinear features used in this study were similarity index, phase synchronization, and nonlinear interdependence. We designed an ANFIS classifier constructed based on these features as input. Fuzzy if-then rules were generated by the ANFIS classifier using the complex relationship of feature space provided during training. The membership function optimization was conducted based on a hybrid learning algorithm. The proposed method achieved highest sensitivity of 80% with false prediction rate as low as 0.46 per hour.

MEDITATION EEG INTERPRETATION BASED ON NOVEL FUZZY-MERGING STRATEGIES AND WAVELET FEATURES

As the advantages of meditation have been outlined literally, scientific exploration of the meditation phenomena becomes significant. Meditation EEG may provide an access to the mental states beyond normal consciousness. It is the first attempt to score the meditation course by EEG. Wavelet analysis and fuzzy c-means (FCM) are applied in the automatic interpretation algorithm. However, FCM applied straightforward to quantitative feature vectors often results in an over-trifling interpretation. As a consequence, this paper presents novel cluster-managing strategies for achieving an interpretation closer to the result of naked-eye examination. The running gray-scale chart, derived by extracting, clustering, and coding the EEG features, reveals five different meditation scenarios differing from those of the controlled subjects.

Electroencephalogram-based pharmacodynamic measures: a review

Pharmacokinetics and pharmacodynamics can provide a useful modeling framework for predicting drug activity and can serve as a basis for dose optimization. Determining a suitable biomarker or surrogate measure of drug effect for pharmacodynamic models can be challenging. The electroencephalograph is a widely-available and non-invasive tool for recording brainwave activity simultaneously from multiple brain regions. Certain drug classes (such as drugs that act on the central nervous system) also generate a reproducible electroencephalogram (EEG) effect. Characterization of such a drug-induced EEG effect can produce pharmacokinetic/pharmacodynamic models useful for titrating drug levels and expediting development of chemically-similar drug analogs. This paper reviews the relevant concepts involved in pharmacokinetic/pharmacodynamic modeling using EEG-based pharmacodynamic measures. In addition, examples of such models for various drugs are organized by drug activity as well as chemical structure and presented.

Patient-specific early seizure detection from scalp electroencephalogram

The objective of this study is to develop a method for automatic detection of seizures before or immediately after clinical onset using features derived from scalp electroencephalogram. This detection method is patient specific. It uses recurrent neural networks and a variety of input features. For each patient, we trained and optimized the detection algorithm for two cases: (1) during the period immediately preceding seizure onset and (2) during the period immediately after seizure onset. Continuous scalp electroencephalogram recordings (duration 15-62 hours, median 25 hours) from 25 patients, including a total of 86 seizures, were used in this study. Preonset detection was successful in 14 of the 25 patients. For these 14 patients, all of the testing seizures were detected before seizure onset with a median preonset time of 51 seconds and false-positive (FP) rate was 0.06/hour. Postonset detection had 100% sensitivity, 0.023/hour FP rate, and median delay of 4 seconds after onset. The unique results of this study relate to preonset detection. Our results suggest that reliable preonset seizure detection may be achievable for a significant subset of patients with epilepsy without use of invasive electrodes.

Automated epilepsy diagnosis using interictal scalp EEG

Over 50 million people worldwide suffer from epilepsy. Traditional diagnosis of epilepsy relies on tedious visual screening by highly trained clinicians from lengthy EEG recording that contains the presence of seizure (ictal) activities. Nowadays, there are many automatic systems that can recognize seizure-related EEG signals to help the diagnosis. However, it is very costly and inconvenient to obtain long-term EEG data with seizure activities, especially in areas short of medical resources. We demonstrate in this paper that we can use the interictal scalp EEG data, which is much easier to collect than the ictal data, to automatically diagnose whether a person is epileptic. In our automated EEG recognition system, we extract three classes of features from the EEG data and build Probabilistic Neural Networks (PNNs) fed with these features. We optimize the feature extraction parameters and combine these PNNs through a voting mechanism. As a result, our system achieves an impressive 94.07% accuracy.

Saccadic spike potentials in gamma-band EEG: characterization, detection and suppression

Analysis of high-frequency (gamma-band) neural activity by means of non-invasive EEG is gaining increasing interest. However, we have recently shown that a saccade-related spike potential (SP) seriously confounds the analysis of EEG induced gamma-band responses (iGBR), as the SP eludes traditional EEG artifact rejection methods. Here we provide a comprehensive profile of the SP and evaluate methods for its detection and suppression, aiming to unveil true cerebral gamma-band activity. The SP appears consistently as a sharp biphasic deflection of about 22 ms starting at the saccade onset, with a frequency band of approximately 20-90 Hz. On the average, larger saccades elicit higher SP amplitudes. The SP amplitude gradually changes from the extra-ocular channels towards posterior sites with the steepest gradients around the eyes, indicating its ocular source. Although the amplitude and the sign of the SP depend on the choice of reference channel, the potential gradients remain the same and non-zero for all references. The scalp topography is modulated almost exclusively by the direction of saccades, with steeper gradients ipsilateral to the saccade target. We discuss how the above characteristics impede attempts to remove these SPs from the EEG by common temporal filtering, choice of different references, or rejection of contaminated trials. We examine the extent to which SPs can be reliably detected without an eye tracker, assess the degree to which scalp current density derivation attenuates the effect of the SP, and propose a tailored ICA procedure for minimizing the effect of the SP.

Bench-to-bedside review: oxygen as a drug

Oxygen is one of the most commonly used therapeutic agents. Injudicious use of oxygen at high partial pressures (hyperoxia) for unproven indications, its known toxic potential, and the acknowledged roles of reactive oxygen species in tissue injury led to skepticism regarding its use. A large body of data indicates that hyperoxia exerts an extensive profile of physiologic and pharmacologic effects that improve tissue oxygenation, exert anti-inflammatory and antibacterial effects, and augment tissue repair mechanisms. These data set the rationale for the use of hyperoxia in a list of clinical conditions characterized by tissue hypoxia, infection, and consequential impaired tissue repair. Data on regional hemodynamic effects of hyperoxia and recent compelling evidence on its anti-inflammatory actions incited a surge of interest in the potential therapeutic effects of hyperoxia in myocardial revascularization and protection, in traumatic and nontraumatic ischemicanoxic brain insults, and in prevention of surgical site infections and in alleviation of septic and nonseptic local and systemic inflammatory responses. Although the margin of safety between effective and potentially toxic doses of oxygen is relatively narrow, the ability to carefully control its dose, meticulous adherence to currently accepted therapeutic protocols, and individually tailored treatment regimens make it a cost-effective safe drug.

A wavelet based analysis of energy redistribution in scalp EEG during epileptic seizures

In this work, wavelet decomposition and multiresolution analysis are used to explore the changes in scalp EEG signals during epileptic seizures. EEG tracings, which include non-epileptic periods, the beginning of seizure and the peak of seizure, have been decomposed in five details using order 10 Daubechies orthonormal wavelets. Energy has then been computed, at each detail, from square wavelet coefficients, in order to unmask the presence of brief episodes of energy relocation among different scales. Results reveal the existence of significant changes in energy distribution at seizure onset; this redistribution, however, exhibits significant differences from one patient to another, and also among different channels in the same patient. Some channels exhibit a significant energy increase at low scales (high frequencies greater than 20 Hz) at seizure onset, whereas energy drops at higher scales. Other channels, however, exhibit energy increase at high scales (frequency less than 15 Hz) revealing a predominance of low-frequency activity. These two patterns may be simultaneously present at seizure onset and may change with different spatial evolution during the subsequent seizure progression. Wavelet analysis appears as a powerful tool for extracting features relative to seizure, and to study their propagation among different regions in the scalp.

Motor Unit Action Potential Duration, II: A New Automatic Measurement Method Based on the Wavelet Transform

The aim of this work is to present and evaluate a new algorithm, based on the wavelet transform, for the automatic measurement of motor unit action potential (MUAP) duration. A total of 240 MUAPs were studied. The waveform of each MUAP was wavelet-transformed, and the start and end points were estimated by regarding the maxima and minima points in a particular scale of the wavelet transform. The results of the new method were compared to the gold standard of duration marker positions obtained by manual measurement. The new method was also compared to a conventional algorithm, which we had found to be best in a previous comparative study. To evaluate the new method against manual measurements, the dispersion of automatic and manual duration markers were analyzed in a set of 19 repeatedly recorded MUAPs. The differences between the new algorithm's marker positions and the gold standard of duration marker positions were smaller than those observed with the conventional method. The dispersion of the new algorithm's marker positions was slightly less than that of the manual one. Our new method for automatic measurement of MUAP duration is more accurate than other available algorithms and more consistent than manual measurements.

Automatic seizure detection based on time-frequency analysis and artificial neural networks

The recording of seizures is of primary interest in the evaluation of epileptic patients. Seizure is the phenomenon of rhythmicity discharge from either a local area or the whole brain and the individual behavior usually lasts from seconds to minutes. Since seizures, in general, occur infrequently and unpredictably, automatic detection of seizures during long-term electroencephalograph (EEG) recordings is highly recommended. As EEG signals are nonstationary, the conventional methods of frequency analysis are not successful for diagnostic purposes. This paper presents a method of analysis of EEG signals, which is based on time-frequency analysis. Initially, selected segments of the EEG signals are analyzed using time-frequency methods and several features are extracted for each segment, representing the energy distribution in the time-frequency plane. Then, those features are used as an input in an artificial neural network (ANN), which provides the final classification of the EEG segments concerning the existence of seizures or not. We used a publicly available dataset in order to evaluate our method and the evaluation results are very promising indicating overall accuracy from 97.72% to 100%.

Time–frequency analysis of spike-wave discharges using a modified wavelet transform

The continuous Morlet wavelet transform was used for the analysis of the time-frequency pattern of spike-wave discharges (SWD) as can be recorded in a genetic animal model of absence epilepsy (rats of the WAG/Rij strain). We developed a new wavelet transform that allows to obtain the time-frequency dynamics of the dominating rhythm during the discharges. SWD were analyzed pre- and post-administration of certain drugs. SWD recorded predrug demonstrate quite uniform time-frequency dynamics of the dominant rhythm. The beginning of the discharge has a short period with the highest frequency value (up to 15 Hz). Then the frequency decreases to 7-9 Hz and frequency modulation occurs during the discharge in this range with a period of 0.5-0.7 s. Specific changes of SWD time-frequency dynamics were found after the administration of psychoactive drugs, addressing different brain mediator and modulator systems. Short multiple SWDs appeared under low (0.5 mg/kg) doses of haloperidol, they are characterized by a fast frequency decrease to 5-6 Hz at the end of every discharge. The frequency of the dominant frequency of SWD was not stable in long lasting SWD after 1.0 mg/kg or more haloperidol: then two periodicities were found. Long lasting SWD seen after the administration of vigabatrin showed a stable frequency of the discharge. The EEG after Ketamin showed a distinct 5 s quasiperiodicity. No clear changes of time-frequency dynamics of SWD were found after perilamine. It can be concluded that the use of the modified Morlet wavelet transform allows to describe significant parameters of the dynamics in the time-frequency domain of the dominant rhythm of SWD that were not previously detected.

Classification of EEG signals using neural network and logistic regression

Epileptic seizures are manifestations of epilepsy. Careful analyses of the electroencephalograph (EEG) records can provide valuable insight and improved understanding of the mechanisms causing epileptic disorders. The detection of epileptiform discharges in the EEG is an important component in the diagnosis of epilepsy. As EEG signals are non-stationary, the conventional method of frequency analysis is not highly successful in diagnostic classification. This paper deals with a novel method of analysis of EEG signals using wavelet transform and classification using artificial neural network (ANN) and logistic regression (LR). Wavelet transform is particularly effective for representing various aspects of non-stationary signals such as trends, discontinuities and repeated patterns where other signal processing approaches fail or are not as effective. Through wavelet decomposition of the EEG records, transient features are accurately captured and localized in both time and frequency context. In epileptic seizure classification we used lifting-based discrete wavelet transform (LBDWT) as a preprocessing method to increase the computational speed. The proposed algorithm reduces the computational load of those algorithms that were based on classical wavelet transform (CWT). In this study, we introduce two fundamentally different approaches for designing classification models (classifiers) the traditional statistical method based on logistic regression and the emerging computationally powerful techniques based on ANN. Logistic regression as well as multilayer perceptron neural network (MLPNN) based classifiers were developed and compared in relation to their accuracy in classification of EEG signals. In these methods we used LBDWT coefficients of EEG signals as an input to classification system with two discrete outputs: epileptic seizure or non-epileptic seizure. By identifying features in the signal we want to provide an automatic system that will support a physician in the diagnosing process. By applying LBDWT in connection with MLPNN, we obtained novel and reliable classifier architecture. The comparisons between the developed classifiers were primarily based on analysis of the receiver operating characteristic (ROC) curves as well as a number of scalar performance measures pertaining to the classification. The MLPNN based classifier outperformed the LR based counterpart. Within the same group, the MLPNN based classifier was more accurate than the LR based classifier.

Forecasting epilepsy from the heart rate signal

Information contained in the R-R interval series, specific to the pre-ictal period, was sought by applying an unsupervised fuzzy clustering algorithm to the N-dimensional phase space of N consecutive interval durations or the absolute value of duration differences. Data sources were individual, complex partial seizures of temporal-lobe epileptics and generalised seizures of rats rendered epileptic with hyperbaric oxygen. Forecasting success was 86% and 82% (zero false positives in resistant rats), respectively, at times ranging from 10 min to 30 s prior to seizure onset Although certain forecasting clusters predominated in the patient group and different ones predominated in the animal group, forecasting on the whole was seizure-specific. The high prediction sensitivity of this method, which matches that of EEG-based methods, seems promising. It is believed that an on-line version of the algorithm, trained on each patient's peri-ictal ECG, could serve as a basis for a simple seizure alarm system.

Seizure detection: evaluation of the Reveal algorithm

OBJECTIVE

The aim of this study is to evaluate an improved seizure detection algorithm and to compare with two other algorithms and human experts.

METHODS

672 seizures from 426 epilepsy patients were examined with the (new) Reveal algorithm which utilizes 3 methods, novel in their application to seizure detection: Matching Pursuit, small neural network-rules and a new connected-object hierarchical clustering algorithm.

RESULTS

Reveal had a sensitivity of 76% with a false positive rate of 0.11/h. Two other algorithms (Sensa and CNet) were tested and had sensitivities of 35.4 and 48.2% and false positive rates of 0.11/h and 0.75/h, respectively.

CONCLUSIONS

This study validates the Reveal algorithm, and shows it to compare favorably with other methods.

SIGNIFICANCE

Improved seizure detection can improve patient care in both the epilepsy monitoring unit and the intensive care unit.

Survey of fuzzy logic applications in brain-related researches

The aim of this study was to survey fuzzy logic (FL) applications in brain researches. In general, these applications are related to pattern recognition for localization in brain structures or tumor detection, image segmentation, and simulations. In recent years, neural networks and FL are gaining popularity. FL is based on the observation of people. The enormous amount of information representation by the brain suggests that FL principles can be useful, especially for complex brain functions. Causal models based on functional neuroanatomy can be then implemented in computer simulations to reflect the dynamical intersection of brain structures. FL is considered as an appropriate tool for modelling and control. FL has been applied in different ways to brain researches. This paper surveys the utilization of FL in brain researches.

A multistage, multimethod approach for automatic detection and classification of epileptiform EEG

An efficient system for detection of epileptic activity in ambulatory electroencephalogram (EEG) must be sensitive to abnormalities while keeping the false-detection rate to a low level. Such requirements could be fulfilled neither by single stage nor by simple method strategy, due to the extreme variety of EEG morphologies and frequency of artifacts. The present study proposes a robust system that combines multiple signal-processing methods in a multistage scheme, integrating adaptive filtering, wavelet transform, artificial neural network, and expert system. The system consists of two main stages: a preliminary screening stage in which data are reduced significantly, followed by an analytical stage. Unlike most systems that merely focus on sharp transients, our system also takes into account slow waves. A nonlinear filter for separation of nonstationary and stationary EEG components is also developed in this paper. The system was evaluated on testing data from 81 patients, totaling more than 800 hours of recordings. 90.0% of the epileptic events were correctly detected. The detection rate of sharp transients was 98.0% and overall false-detection rate was 6.1%. We conclude that our system has good performance in detecting epileptiform activities and the multistage multimethod approach is an appropriate way of solving this problem.

Prediction of epileptic seizures

For almost 40 years, neuroscientists thought that epileptic seizures began abruptly, just a few seconds before clinical attacks. There is now mounting evidence that seizures develop minutes to hours before clinical onset. This change in thinking is based on quantitative studies of long digital intracranial electroencephalographic (EEG) recordings from patients being evaluated for epilepsy surgery. Evidence that seizures can be predicted is spread over diverse sources in medical, engineering, and patent publications. Techniques used to forecast seizures include frequency-based methods, statistical analysis of EEG signals, non-linear dynamics (chaos), and intelligent engineered systems. Advances in seizure prediction promise to give rise to implantable devices able to warn of impending seizures and to trigger therapy to prevent clinical epileptic attacks. Treatments such as electrical stimulation or focal drug infusion could be given on demand and might eliminate side-effects in some patients taking antiepileptic drugs long term. Whether closed-loop seizure-prediction and treatment devices will have the profound clinical effect of their cardiological predecessors will depend on our ability to perfect these techniques. Their clinical efficacy must be validated in large-scale, prospective, controlled trials.

Wavelet transforms of TM joint vibrations: a feature extraction tool for detecting reducing displaced disks

The objective of this study was to determine if wavelet transforms (WTs) of vibrations recorded from temporomandibular joints (TMJs) with reducing displaced disks could be visually separated from WTs of vibrations recorded from normal TM joints by blinded observers. From a continuous series of 124 diagnosed TMD patients, 28 were confirmed with at least one reducing displaced disk. Vibrations were recorded from each affected joint, together with incisal point movements, using BioPAK (BioResearch, Inc., Milwaukee, WI) during opening, closing, and lateral excursions. Identical recordings were taken from 28 patients who were determined to have normal "nondisplacing, nondisplaced" joints. A 3x7 Biorthogonal Spline Wavelet Transform was used to create three-dimensional time-frequency graphs of the vibration events for each subject. Printed copies of the graphs were then shown sequentially to seven blinded observers who were asked to separate them into two groups without any knowledge of their significance. Each observer was independently able to separate the two groups without committing more than one error. We conclude that the vibrations generated by reducing displaced disks are sufficiently different from the vibrations of normal joints to be separable by visual inspection of their respective wavelet transforms.

Automatic segmentation of non-enhancing brain tumors in magnetic resonance images

Tumor segmentation from magnetic resonance (MR) images may aid in tumor treatment by tracking the progress of tumor growth and/or shrinkage. In this paper we present the first automatic segmentation method which separates non-enhancing brain tumors from healthy tissues in MR images to aid in the task of tracking tumor size over time. The MR feature images used for the segmentation consist of three weighted images (T1, T2 and proton density (PD)) for each axial slice through the head. An initial segmentation is computed using an unsupervised fuzzy clustering algorithm. Then, integrated domain knowledge and image processing techniques contribute to the final tumor segmentation. They are applied under the control of a knowledge-based system. The system knowledge was acquired by training on two patient volumes (14 images). Testing has shown successful tumor segmentations on four patient volumes (31 images). Our results show that we detected all six non-enhancing brain tumors, located tumor tissue in 35 of the 36 ground truth (radiologist labeled) slices containing tumor and successfully separated tumor regions from physically connected CSF regions in all the nine slices. Quantitative measurements are promising as correspondence ratios between ground truth and segmented tumor regions ranged between 0.368 and 0.871 per volume, with percent match ranging between 0.530 and 0.909 per volume.

A survey of fuzzy logic monitoring and control utilisation in medicine

Intelligent systems have appeared in many technical areas, such as consumer electronics, robotics and industrial control systems. Many of these intelligent systems are based on fuzzy control strategies which describe complex systems mathematical models in terms of linguistic rules. Since the 1980s new techniques have appeared from which fuzzy logic has been applied extensively in medical systems. The justification for such intelligent systems driven solutions is that biological systems are so complex that the development of computerised systems within such environments is not always a straightforward exercise. In practice, a precise model may not exist for biological systems or it may be too difficult to model. In most cases fuzzy logic is considered to be an ideal tool as human minds work from approximate data, extract meaningful information and produce crisp solutions. This paper surveys the utilisation of fuzzy logic control and monitoring in medical sciences with an analysis of its possible future penetration.