Unified probabilistic framework for predicting and detecting seizure onsets in the brain and multitherapeutic device

• US 6,678,548 B1
• Filed: 10/20/2000
• Issued: 01/13/2004
• Est. Priority Date: 10/20/2000
• Status: Expired due to Term

• Current Assignee: Trustees Of The University Of Pennsylvania (University of Pennsylvania)
• Original Assignee: Trustees Of The University Of Pennsylvania (University of Pennsylvania)
• Inventors: Litt, Brian, Vachtsevanos, George John, Echauz, Javier Ramón, Esteller, Rosana
• Primary Examiner(s): ROBINSON, DANIEL LEON
• Application Number:
  US09/693,423
• Time in Patent Office: 1,180 Days
• Field of Search: 600/544, 600/545, 600/300, 600/378, 340/573, 601/15, 607/45, 607/2
• US Class Current: 600/544
• CPC Class Codes: A61B 5/076 Permanent implantations telemetry using implanted circuitry A61B5/0031; implanted stimulators for therapy A61N1/3605, A61N1/362, A61N1/372
• CPC Class Codes: A61B 5/369 Electroencephalography [EEG] devices for psychotechnics A61B5/16
• CPC Class Codes: A61B 5/37 Intracranial electroencephalography [IC-EEG], e.g. electrocorticography [ECoG]
• CPC Class Codes: A61B 5/372 Analysis of electroencephalograms
• CPC Class Codes: A61B 5/4094 Diagnosing or monitoring seizure diseases, e.g. epilepsy
• CPC Class Codes: A61B 5/7267 involving training the classification device
• CPC Class Codes: A61B 5/7275 Determining trends in physiological measurement data; Predicting development of a medical condition based on physiological measurements, e.g. determining a risk factor
• CPC Class Codes: A61N 1/36017 with leads or electrodes penetrating the skin
• CPC Class Codes: A61N 1/36064 Epilepsy
• CPC Class Codes: A61N 1/36082 Cognitive or psychiatric applications, e.g. dementia or Alzheimer's disease
• CPC Class Codes: A61N 1/36135 using physiological parameters
• CPC Class Codes: G06F 18/217 Validation; Performance evaluation; Active pattern learning techniques
• CPC Class Codes: G06F 18/2413 based on distances to training or reference patterns
• CPC Class Codes: G06F 18/24155 Bayesian classification
• CPC Class Codes: G16H 20/30 relating to physical therapies or activities, e.g. physiotherapy, acupressure or exercising
• CPC Class Codes: G16H 50/20 for computer-aided diagnosis, e.g. based on medical expert systems
• CPC Class Codes: G16H 50/30 for calculating health indices; for individual health risk assessment
• CPC Class Codes: G16H 50/50 for simulation or modelling of medical disorders

Abstract

A method and an apparatus for predicting and detecting epileptic seizure onsets within a unified multiresolution probabilistic framework, enabling a portion of the device to automatically deliver a progression of multiple therapies, ranging from benign to aggressive as the probabilities of seizure warrant. Based on novel computational intelligence algorithms, a realistic posterior probability function P(S_t|x) representing the probability of one or more seizures starting within the next T minutes, given observations x derived from IEEG or other signals, is periodically synthesized for a plurality of prediction time horizons. When coupled with optimally determined thresholds for alarm or therapy activation, probabilities defined in this manner provide anticipatory time-localization of events in a synergistic logarithmic-like array of time resolutions, thus effectively circumventing the performance vs. prediction-horizon tradeoff of single-resolution systems. The longer and shorter prediction time scales are made to correspond to benign and aggressive therapies respectively. The imminence of seizure events serves to modulate the dosage and other parameters of treatment during open-loop or feedback control of seizures once activation is triggered. Fast seizure onset detection is unified within the framework as a degenerate form of prediction at the shortest, or even negative, time horizon. The device is required to learn in order to find the probabilistic prediction and control strategies that will increase the patient'"'"'s quality of life over time. A quality-of-life index (QOLI) is used as an overall guide in the optimization of patient-specific signal features, the multitherapy activation decision logic, and to document if patients are actually improving.

Claims

1. A method for automatically predicting and preventing the electrographic onset of a seizure in an individual, comprising the acts of:

monitoring a plurality of signals indicative of the activity of the brain of the individual;

extracting a set of features from the signals and forming an optimal feature vector;

synthesizing a probability vector based on the optimal feature vector as an estimator of the likelihood of seizure for a plurality of prediction time intervals; and

preventing the electrographic onset of a seizure by the automatic application of at least one intervention measure that is commensurate with the likelihood of seizure.

2. The method for automatically predicting and preventing the electrographic onset of a seizure of claim 1 further comprising the act of displaying the probability vector of oncoming seizures for a plurality of prediction time intervals to indicate both the time frame when the electrographic onsets are expected to occur and a degree of confidence in the predictions of electrographic onsets.

3. The method for automatically predicting and preventing the electrographic onset of a seizure of claim 1 wherein the act of preventing the electrographic onset of seizures automatically applies the minimally required intervention measure thereby minimizing associated side effects.

4. The method for automatically predicting and preventing the electrographic onset of a seizure of claim 3 wherein the minimally required intervention measure starts with a benign form of therapy and continues with more aggressive therapies as the probability vector continuously changes with decreasing time to electrographic onset.

5. The method for automatically predicting and preventing the electrographic onset of a seizure of claim 3 wherein the minimally required intervention measure couples minimally invasive benign therapies to long prediction time intervals, aggressive therapies to short prediction time intervals and gradually varying moderate forms of therapy to prediction time intervals that are between the long and short prediction time intervals.

6. The method for automatically predicting and preventing the electrographic onset of a seizure of claim 5 wherein the minimally required intervention measure applied matches the tolerance for false positive and false negative prediction errors for each prediction time interval.

7. The method for automatically predicting and preventing the electrographic onset of a seizure of claim 1 wherein the intervention measure includes at least one of cognitive stimulation, sensory stimulation, biofeedback, electrical stimulation and pharmacological infusion.

8. The method for automatically predicting and preventing the electrographic onset of a seizure of claim 1 further comprising the act of applying a closed loop feedback control law and commanding a therapy actuator in order to regulate the seizure probability vector.

9. The method for automatically predicting and preventing the electrographic onset of a seizure of claim 8 wherein the applied closed loop feedback control law and therapy actuator includes any of a pole-placed, state-variable feedback, optimal linear-quadratic, gain scheduled, multilevel bang-bang, model-predictive, and intelligent control strategies to regulate the seizure probability vector as a controlled variable.

10. The method for automatically predicting and preventing the electrographic onset of a seizure of claim 6 wherein the intervention measure is triggered by prediction thresholds with high sensitivity and low specificity.

11. The method for automatically predicting and preventing the electrographic onset of a seizure of claim 1 wherein the intervention measure is delivered in at least one of a region of onset and a distribution region surrounding the region of onset.

12. The method for automatically predicting and preventing the electrographic onset of a seizure of claim 1 wherein the intervention measure is delivered in subcortical regions including at least one of the thalamus, basal ganglia, and other deep nuclei.

13. The method for automatically predicting and preventing the electrographic onset of a seizure of claim 1 wherein if the electrographic onset occurs, applying treatment to at least one of a general region of onset and deep brain structures to modulate the behavior of the seizure focus.

14. The method for automatically predicting and preventing the electrographic onset of a seizure of claim 1 wherein the intervention measure application includes at least one of:

rhythmic electrical pacing that changes in frequency, intensity and distribution as the probability of a seizure onset reaches and exceeds a threshold;

chaos control pacing;

random electrical stimulation to interfere with developing coherence in activity in a region of, and surrounding, an epileptic focus;

depolarization or hyperpolarization stimuli to silence or suppress activity in actively discharging regions, or regions at risk for seizure spread.

15. The method for automatically predicting and preventing the electrographic onset of a seizure of claim 14 wherein the intervention measure is delivered to a plurality of electrodes to provide a surround inhibition to prevent a progression of a seizure precursor.

16. The method for automatically predicting and preventing the electrographic onset of a seizure of claim 14 wherein the intervention measure is delivered sequentially in a wave that covers a cortical or subcortical region of tissue so as to progressively inhibit normal or pathological neuronal function in the covered region.

17. The method for automatically predicting and preventing the electrographic onset of a seizure of claim 1 wherein the intervention measure application is an infusion of a therapeutic chemical agent into a brain region where seizures are generated, or to which they may spread.

18. The method for automatically predicting and preventing the electrographic onset of a seizure of claim 17 wherein the chemical agent is delivered in greater quantity, concentration or spatial distribution as the probability of seizure increases.

19. The method for automatically predicting and preventing the electrographie onset of a seizure of claim 17 wherein the intervention measure is applied to at least one of an epilectic focus, an area surrounding the epilectic focus, a region involved in an early spread, and a central or deep brain region to modulate seizure propagation.

20. The method for automatically predicting and preventing the electrographic onset of a seizure of claim 17 wherein the therapeutic chemical agent is activated by oxidative stress and increases in concentration and distribution as the probability of seizure increases.

21. The method for automatically predicting and preventing the electrographic onset of a seizure of claim 1 wherein the intervention measure is delivered to central nerves or blood vessels in a graduated manner as the probability of seizure increases.

22. The method for automatically predicting and preventing the electrographic onset of a seizure of claim 1 wherein the intervention measure is a plurality of artificial neuronal signals delivered to disrupt eletrochemical traffic on at least one neuronal network that includes or communicates with an ictal onset zone.