Burst suppression monitor for induced coma

• US 8,649,855 B2
• Filed: 10/01/2007
• Issued: 02/11/2014
• Est. Priority Date: 09/29/2006
• Status: Expired due to Fees

• Current Assignee: Regents of the University of California (University of California), Wavestate, Inc.
• Original Assignee: Regents of the University of California (University of California), Wavestate, Inc.
• Inventors: Kaiser, David A., Bier, Michael
• Primary Examiner(s): Natnithithadha, Navin
• Assistant Examiner(s): MESSERSMITH, ERIC J
• Application Number:
  US11/865,727
• Publication Number:
  US 20080249431A1
• Time in Patent Office: 2,325 Days
• Field of Search: 600544-545
• US Class Current: 600/544
• CPC Class Codes: A61B 5/316 Modalities, i.e. specific diagnostic methods
• CPC Class Codes: A61B 5/372 Analysis of electroencephalograms
• CPC Class Codes: A61B 5/4821 Determining level or depth of anaesthesia A61B5/1106 takes precedence
• CPC Class Codes: A61B 5/4839 combined with drug delivery
• CPC Class Codes: A61B 5/7264 Classification of physiological signals or data, e.g. using neural networks, statistical classifiers, expert systems or fuzzy systems neural networks per se G06N3/00; expert systems per se G06N5/00

Abstract

Methods and systems for monitoring subjects, analyzing EEG data, and maintaining subjects at desired sedation states.

Claims

1. A method of monitoring a sedated subject using a logic processor in a monitoring system comprising:

receiving digital data representing said subject'"'"'s EEG signal at said processor;

using said processor to determine one or more burst onsets;

using said processor to determine one or more suppression onsets;

using said processor to determine one or more interburst or burst suppression intervals by comparing one or more burst onsets to one or more previous suppression onsets;

using said processor to determine a true interburst interval or true mean burst suppression (BS) interval for a specified time length with statistical confidence wherein said mean burst suppression interval indicates time between burst onsets and previous suppression onsets, regardless of burst duration;

using said mean burst suppression interval to inform therapy for said subject to maintain said desired state of sedation;

calculating an absolute amplitude difference across said EEG signal, with an absolute amplitude difference history interval;

smoothing said absolute amplitude difference to minimizes effects of signal variation on onset detection; and

detecting burst onset when an absolute smoothed amplitude difference exceeds a median absolute amplitude times a specified multiplication factor.

2. The method of claim 1 further wherein:

said using said mean burst suppression interval to inform therapy comprising presenting data indicating said mean burst suppression interval with statistical confidence to a user, thereby providing the user information to guide adjusting therapy to a subject to maintain the desired state of sedation.

3. The method of claim 1 further wherein:

said using said mean burst suppression interval to indicate therapy comprising using a processor controlled medical device to adjust administration of one or more therapies using said mean burst suppression (BS) interval.

4. The method of claim 1 further wherein:

said subject is being maintained in an induced coma and said method is used by an automated system to prevent over-sedation or under-sedation.

5. The method of claim 1 further comprising:

using said processor to perform statistical analysis to correct for noise variability, other isoelectric variability, or other electric disturbances in said EEG.

6. The method of claim 1 further comprising:

presenting data to a user indicating a true interburst interval with statistical accuracy.

7. The method of claim 1 further wherein:

said digital data representing said subject'"'"'s EEG is a stream of digital samples of amplitude values at different data capture time points;

and determining burst or suppression onset further by a method comprising;

computing changes in amplitude substantially at each data capture point;

using said processor to perform a smoothing operation at each data capture point; and

using said processor to perform threshold differentiation at each data capture point.

8. The method of claim 7 further comprising:

using said processor to determine a possibility of an onset of burst or suppression for substantially each and every digital sample.

9. The method of claim 1 further wherein said method is embodied as an automatic computer-aided segmentation algorithm to improve patient care by guiding titration of sedatives.

10. The method of claim 1 further comprising:

using said processor to pre-process said EEG signal data to perform noise reduction by mathematical smoothing operations such as smoothing and filtering; and

further wherein said noise reduction comprising one or more of;

using a 3-sample-skip comparison wherein amplitude at a time 0 is compared to amplitude at a time 3, amplitude at time 1 to amplitude at time 4, etc.;

using a 1-sample skip to capture rare low-amplitude bursts in some patients;

using a 2-sample skip as a default case.

11. The method of claim 1 further comprising:

using data from one or more EEG channels to detect variance in BS generation across scalp recordings to strengthen reliability of detection of burst onset and suppression onset.

12. The method of claim 1 further comprising:

calculating absolute amplitude difference (d_i) in microvolts continuously across said EEG signal, with a 1 second history;

using smoothing to minimizes effects of signal variation on onset detection; and

detecting burst onset when an absolute amplitude difference (d_i) exceeds 2 times a median (m) absolute amplitude.

13. The method of claim 12 further comprising:

calculating said absolute amplitude difference (d_i) according to

14. The method of claim 1 further comprising:

calculating absolute amplitude difference (d_i) in microvolts continuously across said EEG signal, with a 1 second history;

using smoothing to minimizes effects of signal variation on onset detection to determine a;

smoothed amplitude difference; and

detecting burst onset when an absolute smoothed amplitude difference exceeds B times a median (m) absolute amplitude;

wherein B is chosen theoretically or empirically and can be adjusted per patient or learned by an algorithm for said subject.

15. The method of claim 12 further comprising:

determining a median absolute amplitude (MAA) from a longer running history;

updating said median absolute amplitude periodically.

16. The method of claim 1 further wherein:

said mean burst suppression interval is expressed in units of seconds.

17. The method of claim 1 further comprising:

using said processor to determine a statistical confidence value (P) of said mean burst suppression interval;

using said mean burst suppression interval along with said statistical confidence value to adjust therapy to a subject to maintain said desired state of sedation.

18. The method of claim 12 further comprising:

ignoring a particular suppression onset whenever a burst onset occurs less than a specified time (e.g., 500 or 1000 ms) after said particular suppression onset.

19. The method of claim 3 further comprising:

determining central tendency of burst onsets over a period of time and to a user presenting results in units of bursts per minute or mean burst interval in seconds or like measure of central tendency.

20. The method of claim 19 further comprising:

allowing a user to set a history over which said bursts per minute is measured;

using a default history duration if a user so indicates or does not select a different history.

21. The method of claim 12 further comprising:

resetting a time count (e.g., number of digitized samples since last burst onset) to zero at each burst onset, wherein a preceding immediately prior time count is designated a current burst suppression (BS) interval.

22. The method of claim 12 further comprising:

said mean burst suppression interval ( b/R) is expressed in units of seconds and is determined by said logic processor;

wherein R=sample rate/sec and b is determined according to;

23. The method of claim 12 further comprising:

indicating confidence at 95% if