	US 11,980,473 B2
	Seizure onset zone localization
Brent M. Berry, Rochester, MN (US); Gary C. Sieck, Rochester, MN (US); Gregory A. Worrell, Rochester, MN (US); Benjamin H. Brinkmann, Byron, MN (US); Yogatheesan Varatharajah, Urbana, IL (US); Vaclav Kremen, Rochester, MN (US); Ravishankar Krishnan Iyer, Champaign, IL (US); Zbigniew Kalbarczyk, Urbana, IL (US); and Jan Cimbalnik, Brno (CZ)
Assigned to Mayo Foundation for Medical Education and Research, Rochester, MN (US); and The Board of Trustees of the University of Illinois and St. Anne's University Hospital Brno, Urbana, IL (US)
Appl. No. 16/616,771
Filed by Brent M. Berry, Rochester, MN (US); Gary C. Sieck, Rochester, MN (US); Gregory A. Worrell, Rochester, MN (US); Benjamin H. Brinkmann, Byron, MN (US); Yogatheesan Varatharajah, Urbana, IL (US); Vaclav Kremen, Rochester, MN (US); Ravishankar Krishnan Iyer, Champaign, IL (US); Zbigniew Kalbarczyk, Urbana, IL (US); and Jan Cimbalnik, Brno (CZ)
PCT Filed May 25, 2018, PCT No. PCT/US2018/034703 § 371(c)(1), (2) Date Nov. 25, 2019, PCT Pub. No. WO2018/218174, PCT Pub. Date Nov. 29, 2018.
Claims priority of provisional application 62/511,239, filed on May 25, 2017.
Prior Publication US 2020/0178832 A1, Jun. 11, 2020
Int. Cl. A61B 5/369 (2021.01); A61B 5/00 (2006.01); A61B 5/07 (2006.01)

CPC A61B 5/369 (2021.01) [A61B 5/076 (2013.01); A61B 5/4094 (2013.01); A61B 5/725 (2013.01)]

19 Claims

1. A computer-implemented method, comprising:

obtaining, by a computing system, and for each of a plurality of sensor channels, a respective set of electroencephalogram (EEG) data for the sensor channel over a first interictal time interval, each sensor channel corresponding to a different one of a plurality of EEG sensors disposed at different locations of a brain of a mammal;

segmenting, by the computing system, and for each of the plurality of sensor channels, the respective set of EEG data for the sensor channel into a plurality of EEG data segments, each EEG data segment corresponding to one of a plurality of sub-intervals of the first interictal time interval;

for each sensor channel of the plurality of sensor channels and each of the plurality of sub-intervals:

(a) generating, based on analysis of the EEG data segment for the sub-interval, a current classification of the sensor channel at the sub-interval as either (i) a non-epileptogenic sensor channel for an EEG sensor that is likely not disposed at or near an epileptogenic region of the brain, or (ii) an epileptogenic sensor channel for an EEG sensor that is likely disposed at or near an epileptogenic region of the brain; and

(b) using the current classification of the sensor channel at the sub-interval to update a Bayesian metric for the sensor channel, the Bayesian metric determined based on the current classification of the sensor channel at the sub-interval and classifications of the sensor channel from any preceding sub-intervals in the first interictal time interval, the Bayesian metric indicating a long-term classification of the sensor channel expressed as a cumulative belief that the sensor channel is either a non-epileptogenic sensor channel or an epileptogenic sensor channel; and

providing, by the computing system, and for each of one or more of the plurality of sensor channels based on the Bayesian metric for the sensor channel, an indication of whether the sensor channel has an EEG sensor that is likely disposed at or near an epileptogenic region of the brain.