Responsive therapy with multi-stage detection of seizures in human epilepsy
Abstract number :
698
Submission category :
3. Neurophysiology / 3E. Brain Stimulation
Year :
2020
Submission ID :
2423038
Source :
www.aesnet.org
Presentation date :
12/7/2020 9:07:12 AM
Published date :
Nov 21, 2020, 02:24 AM
Authors :
Vladimir Sladky, Mayo Clinic; International Clinical Research Center - St. Anne's University Hospital Brno; Vaclav Kremen - Mayo Clinic; Filip Mivalt - Mayo Clinic, Brno University of Technology; Petr Nejedly - Mayo Clinic; Benjamin Brinkmann - Mayo Clinic; Nisali Gunawardane - Yale University School of Medicine; Richard Luo - Dartmouth-Hitchcock Medical Center; Bogdan Litvinov - Yale University School of Medicine; Courtney Yotter - Yale University School of Medicine; Barbara Jobst - Dartmouth-Hitchcock Medical Center; Gregory Worrell - Mayo Clinic; Hal Blumenfeld - Yale University School of Medicine
Rationale:
Electrical brain stimulation (EBS) has been used successfully for treatment of drug resistant temporal lobe epilepsy. In particular, responsive neural stimulation (RNS) is an approach that delivers a brief electrical stimulation burst when epileptiform activity or seizure is detected. The RNS therapy is designed to be delivered to the brain region generating seizures. In the future it might be beneficial to track whether the initial stimulation therapy terminated the seizure, and in cases where stimulation has not stopped the seizure the therapy can be switched to a different network target. In that case, cascade detectors need to be enabled within the implanted neural stimulator (INS). We modelled, implemented, and investigated a multi-stage therapy control approach embedded in an INS to stimulate multiple brain targets with different stimulation settings.
Method:
We created a database of intracranially recorded hippocampal seizures and interictal periods in patients undergoing surgical evaluation for drug resistant temporal lobe epilepsy (Yale, Dartmouth, Mayo Clinic). We developed a bench-top system that enables replaying seizures in a saline tank. We used the system to test detectors embedded in the Medtronic Investigational Summit RC+S ™. We tested a scenario with two coupled embedded detectors. The first detector was used to detect the hippocampal seizures and initiate electrical stimulation. When seizure was not interrupted after a specified duration the second detector, in the presence of ongoing hippocampal stimulation, switched the stimulation to a different target with different stimulation parameters. Here we tested the approach using patient specific detector parameters.
Results:
We tested the embedded two-stage detectors on data collected from 6 patients (200 seizures and 6 interictal days). We achieved 100 % sensitivity and specificity 97 % in the first stage patient specific detector. The average lag from seizure onset to detection was 3.3 seconds. The second detector had 96 % sensitivity with 99.5 % specificity, and 10.62 seconds average lag from seizure onset.
Conclusion:
Our results indicate that embedded multi-stage detection is possible in INS devices. We hypothesize that optimized, multi-site stimulation might prove useful within next-generation EBS devices that target epilepsy networks.
Funding:
:NIH/NINDS Brain Initiative UG3NS112826
Neurophysiology