Seizure warning and prediction

Details Seizure warning and prediction Seizure warning and prediction

1999-09-22

2001-10-16

Lacyk, John P. (Department: 3736)

Surgery

Diagnostic testing

Detecting brain electric signal

Reexamination Certificate

active

06304775

ABSTRACT:

The research and development effort associated with the subject matter of this patent application was support by the National Institutes of Health under grant no. R01 NS31451.
FIELD OF INVENTION
The present invention involves the field of signal processing. More particularly, the present invention involves the processing of electrical and/or electromagnetic signals generated by the brain.
BACKGROUND
Epilepsy is a chronic disorder characterized by recurrent brain dysfunction caused by paroxysmal electrical discharges in the cerebral cortex. If untreated, an individual afflicted with epilepsy is likely to experience repeated seizures, which typically involve some level of impaired consciousness. Some forms of epilepsy can be successfully treated through medical therapy. However, medical therapy is less effective with other forms of epilepsy, including Temporal Lobe Epilepsy (TLE) and Frontal Lobe Epilepsy (FLE). With TLE and FLE, removing the portion of the hippocampus and/or cerebral cortex responsible for initiating the paroxysmal electrical discharges, known as the epileptogenic focus, is sometimes performed in an effort to control the seizures.
For quite some time, a few investigators in the medical research community have attempted to develop techniques which effectively provide seizure prediction and/or seizure warning, where seizure prediction will be understood to involve long-range forecasting of seizure-onset time, whereas seizure warning will be understood to involve long-range indications of conditions conducive to an impending seizure. As one skilled in the art will surely appreciate, any such technique would have numerous clinical as well as non-clinical applications. For example, in order to more effectively treat patients that are resistant to conventional medical therapy, such a technique might be used in conjunction with a device, perhaps an implanted device, designed to deliver a dosage of anti-seizure medication into the patient's bloodstream for the purpose of averting an impending seizure.
In another example, such a technique could be used during pre-surgical evaluations to assist in pinpointing the epileptogenic focus, which is to be removed during surgery. It is understood that during a seizure, blood flow to the epileptogenic focus significantly increases. If certain radio-labeled ligands are injected into the patient's bloodstream in a timely manner, it is possible to monitor that increased blood flow using radiography, thereby allowing a physician to accurately pinpoint the boundaries of the epileptogenic focus. A true seizure prediction and/or warning technique would, ideally, provide an indication of an impending seizure well in advance so as to provide sufficient time to prepare for and administer, for example, the aforementioned radiography ligand.
The most important tool for evaluating the physiological states of the brain is the electroencephalogram (EEG). The standard for analysis and interpretation of the EEG is visual inspection of the graphic tracing of the EEG by a trained clinical electroencephalographer. There is no established method for predicting seizure onset by visual analysis of the EEG. Traditional signal processing techniques yield little practical information about the EEG signal. Such methods, however, are limited in their effectiveness because the brain is a multidimensional system that produces nonlinear signals and exhibits spatial as well as temporal properties.
Moreover, signal processing techniques that simply employ standard, linear, time series analysis methods cannot possibly detect the spatio-temporal properties that are critical in providing effective seizure warning.
To date, there are no known linear or non-linear techniques capable of providing seizure detection, warning, or the like, in advance of seizure onset. At best, present techniques provide, with less than desirable accuracy, seizure detection during the very early stages of a seizure discharge in the EEG (i.e., a few seconds after the initial discharge). The onset of the seizure discharge in the EEG may precede the clinical manifestations (e.g., behavioral and neuromotor responses) of the seizure by up to several seconds, particularly where intra-cranial electrodes are employed for EEG recordings. Because the EEG manifestation may be detected just a few seconds prior to the clinical manifestations of the seizure, some investigators have claimed the ability to predict seizures through evaluation of the EEG. Osorio et al., “Real-time Automated Detection and Quantitative Analysis of Seizures and Short-term Prediction of Clinical Onset,” Epilepsia, vol. 39, pp. 615-627, 1998. Such claims are misleading since these techniques simply detect the EEG manifestation of the seizure. These techniques do not, however, provide seizure prediction. Unfortunately, the few seconds afforded by these early seizure detection techniques are insufficient to support practical applications such as the medical intervention therapy and in-patient applications mentioned previously. For example, with respect to employing a seizure detection/warning technique for medical intervention therapy, seizure detection/warnings that precede seizure onset by 5, 10 or even 60 seconds are unlikely to offer any benefit because any medication administered at that time would not have time to reach a sufficient brain concentration to prevent an impending seizure. Even techniques that may be capable of detecting and/or generating seizure warnings no more than a few minutes prior to seizure onset may not support such a treatment. Also, with regard to utilizing a seizure detection/warning technique to support various in-patient applications, present techniques cannot consistently and accurately provide the timely seizure detection/warnings needed to alert medical staff members so they can properly observe the impending seizure, administer medication to avoid the seizure, or prepare for and perform any pre-surgical procedures such as the radiography procedure described above.
A few of the more recently developed techniques have gone beyond simple, linear, time series analysis in an attempt to provide more timely and more accurate seizure detection/warning capabilities. In accordance with one of these most recent techniques, signals from one or more EEG channels are sampled over relatively short time intervals. A high dimensional state space plot is then generated from each channel, using a common technique called the “Method of Delays.” A more detailed explanation of the Method of Delays can be found in L. lasemidis et al., “Quantification of Hidden Time Dependencies in the EEG within the Framework of Nonlinear Dynamics”, in Nonlinear Dynamical Analysis of the EEG, eds. B. H. Jansen & M. E. Brandt (World Scientific, Singapore, 1993), pp. 30-47. See also, F. Takens, “Detecting Strange Attractors in Turbulence”, in Lec. Notes Math., eds. D. A. Rand & L. S. Young (Springer-Verlag, 1980), 898, pp. 366-381; and H. Whitney, “Differentiable Manifolds”, Ann. Math., 37 (1936), pp. 645-680. Each state space plot, in turn, is used to derive correlation integrals for the corresponding signal, where the correlation integrals reflect the complexity (e.g., the correlation dimension, predictability indices) associated with the corresponding signal. A significant drop in the correlation integral values or the correlation dimension (i.e., a reduction in complexity) over time at specific brain sites can be used to trigger an impending seizure warning. See J. Martinerie et al., “Epileptic Seizures can be Anticipated by Nonlinear Analysis”, Nature Medicine, vol. 4, pp. 1173-1176, 1998. See also, C. E. Elger et al., “Seizure Prediction by Nonlinear Time Series Analysis of Brain Electrical Activity,” European Journal of Neuroscience, vol. 10, pp. 786-789, 1998.
There are numerous deficiencies associated with the above-identified technique. For instance, the estimated measure of signal complexity is unreliable, as it depends on the brain state, and segments without epileptiform activity are arbitrarily selected as refe

Affiliated with

Also associated with

Rating

Seizure warning and prediction does not yet have a rating. At this time, there are no reviews or comments for this patent.

If you have personal experience with Seizure warning and prediction, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Seizure warning and prediction will most certainly appreciate the feedback.

Rate now

Comments { 0 }

Profile ID: LFUS-PAI-O-2588899