Back to results
US 2013/0096839 A1Published application
Seizure detection methods, apparatus, and systems using a wavelet transform maximum modulus algorithm
AssigneeFlint Hills Scientific, L.L.C.
InventorsIvan Osorio, Alexey Lyubushin, Didier Sornette
Life of this patent
Estimating the term…
Abstract
Methods for detecting a seizure, by use of a wavelet transform maximum modulus (WTMM) algorithm applied to body data. A non-transitive, computer-readable storage device for storing data that when executed by a processor, perform such a method.
Claims23
A non-transitory computer readable program storage unit encoded with instructions that, when executed by a computer, perform a method, comprising:
- receiving a time series of a first body signal of the patient,
- determining at least two sliding adjacent time windows for said time series of said first body signal comprising a foreground window and a background window;
- extracting at least one first long chain from said foreground window and at least one second long chain from said background window;
- applying a continuous wavelet transform to each of said long chains, to yield a wavelet transform maximum modulus (WTMM) skeleton for each window;
- determining a first variance of the WTMM skeleton from said foreground window and a second variance of the WTMM skeleton from said background window;
- generating a stepwise approximation of changes in said variance;
- determining a seizure onset in response to a determination that an output of the stepwise approximation of changes in said first variance has reached an onset threshold;
- and
- determining a seizure termination in response to a determination that an output of the stepwise approximation of changes in said second variance has reached a termination threshold.
6 more elements
A method for detecting an onset and a termination of a seizure from a patient body signal, comprising:
- receiving a time series of a first body signal of the patient,
- determining at least two sliding adjacent time windows for said time series of said first body signal comprising a foreground window and a background window;
- applying a wavelet transform maximum modulus (WTMM) skeleton to at least one long chain extracted from said windowed signals using a continuous wavelet transform;
- determining a variance of the output of the previous step based upon the WTMM skeleton;
- generating a stepwise approximation of the changes in variance of the windowed signal as analyzed using previous steps;
- determining a seizure onset in response to a determination that an output of the stepwise approximation of the changes in variance in the foreground window has reached an onset threshold;
- and
- determining a seizure termination in response to a determination that an output of the stepwise approximation of the changes in variance in the background window has reached a termination threshold.
5 more elements
Dependencies are inferred from the claim text.