A Comparison of Energy-Efficient Seizure Detectors for Implantable Neurostimulation Devices
IntroductionAbout 30% of epilepsy patients are resistant to treatment with antiepileptic drugs, and only a minority of these are surgical candidates. A recent therapeutic approach is the application of electrical stimulation in the early phases of a seizure to interrupt its spread across the brain....
| Main Authors: | , , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Frontiers Media S.A.
2022-03-01
|
| Series: | Frontiers in Neurology |
| Subjects: | |
| Online Access: | https://www.frontiersin.org/articles/10.3389/fneur.2021.703797/full |
| _version_ | 1818910064233152512 |
|---|---|
| author | Farrokh Manzouri Marc Zöllin Simon Schillinger Simon Schillinger Matthias Dümpelmann Ralf Mikut Peter Woias Laura Maria Comella Andreas Schulze-Bonhage Andreas Schulze-Bonhage |
| author_facet | Farrokh Manzouri Marc Zöllin Simon Schillinger Simon Schillinger Matthias Dümpelmann Ralf Mikut Peter Woias Laura Maria Comella Andreas Schulze-Bonhage Andreas Schulze-Bonhage |
| author_sort | Farrokh Manzouri |
| collection | DOAJ |
| description | IntroductionAbout 30% of epilepsy patients are resistant to treatment with antiepileptic drugs, and only a minority of these are surgical candidates. A recent therapeutic approach is the application of electrical stimulation in the early phases of a seizure to interrupt its spread across the brain. To accomplish this, energy-efficient seizure detectors are required that are able to detect a seizure in its early stages.MethodsThree patient-specific, energy-efficient seizure detectors are proposed in this study: (i) random forest (RF); (ii) long short-term memory (LSTM) recurrent neural network (RNN); and (iii) convolutional neural network (CNN). Performance evaluation was based on EEG data (n = 40 patients) derived from a selected set of surface EEG electrodes, which mimic the electrode layout of an implantable neurostimulation system. As for the RF input, 16 features in the time- and frequency-domains were selected. Raw EEG data were used for both CNN and RNN. Energy consumption was estimated by a platform-independent model based on the number of arithmetic operations (AOs) and memory accesses (MAs). To validate the estimated energy consumption, the RNN classifier was implemented on an ultra-low-power microcontroller.ResultsThe RNN seizure detector achieved a slightly better level of performance, with a median area under the precision-recall curve score of 0.49, compared to 0.47 for CNN and 0.46 for RF. In terms of energy consumption, RF was the most efficient algorithm, with a total of 67k AOs and 67k MAs per classification. This was followed by CNN (488k AOs and 963k MAs) and RNN (772k AOs and 978k MAs), whereby MAs contributed more to total energy consumption. Measurements derived from the hardware implementation of the RNN algorithm demonstrated a significant correlation between estimations and actual measurements.DiscussionAll three proposed seizure detection algorithms were shown to be suitable for application in implantable devices. The applied methodology for a platform-independent energy estimation was proven to be accurate by way of hardware implementation of the RNN algorithm. These findings show that seizure detection can be achieved using just a few channels with limited spatial distribution. The methodology proposed in this study can therefore be applied when designing new models for responsive neurostimulation. |
| first_indexed | 2024-12-19T22:36:52Z |
| format | Article |
| id | doaj.art-88c00820394049fcb47253ff42478faf |
| institution | Directory Open Access Journal |
| issn | 1664-2295 |
| language | English |
| last_indexed | 2024-12-19T22:36:52Z |
| publishDate | 2022-03-01 |
| publisher | Frontiers Media S.A. |
| record_format | Article |
| series | Frontiers in Neurology |
| spelling | doaj.art-88c00820394049fcb47253ff42478faf2022-12-21T20:03:11ZengFrontiers Media S.A.Frontiers in Neurology1664-22952022-03-011210.3389/fneur.2021.703797703797A Comparison of Energy-Efficient Seizure Detectors for Implantable Neurostimulation DevicesFarrokh Manzouri0Marc Zöllin1Simon Schillinger2Simon Schillinger3Matthias Dümpelmann4Ralf Mikut5Peter Woias6Laura Maria Comella7Andreas Schulze-Bonhage8Andreas Schulze-Bonhage9Epilepsy Center, Department of Neurosurgery, Medical Center — University of Freiburg, Faculty of Medicine, University of Freiburg, Freiburg, GermanyLaboratory for Design of Microsystems, Department of Microsystems Engineering — IMTEK, University of Freiburg, Freiburg, GermanyLaboratory for Design of Microsystems, Department of Microsystems Engineering — IMTEK, University of Freiburg, Freiburg, GermanyInstitute for Automation and Applied Informatics, Karlsruhe Institute of Technology, Eggenstein-Leopoldshafen, GermanyEpilepsy Center, Department of Neurosurgery, Medical Center — University of Freiburg, Faculty of Medicine, University of Freiburg, Freiburg, GermanyInstitute for Automation and Applied Informatics, Karlsruhe Institute of Technology, Eggenstein-Leopoldshafen, GermanyLaboratory for Design of Microsystems, Department of Microsystems Engineering — IMTEK, University of Freiburg, Freiburg, GermanyLaboratory for Design of Microsystems, Department of Microsystems Engineering — IMTEK, University of Freiburg, Freiburg, GermanyEpilepsy Center, Department of Neurosurgery, Medical Center — University of Freiburg, Faculty of Medicine, University of Freiburg, Freiburg, GermanyFaculty of Medicine, Center for Basics in NeuroModulation, University of Freiburg, Freiburg, GermanyIntroductionAbout 30% of epilepsy patients are resistant to treatment with antiepileptic drugs, and only a minority of these are surgical candidates. A recent therapeutic approach is the application of electrical stimulation in the early phases of a seizure to interrupt its spread across the brain. To accomplish this, energy-efficient seizure detectors are required that are able to detect a seizure in its early stages.MethodsThree patient-specific, energy-efficient seizure detectors are proposed in this study: (i) random forest (RF); (ii) long short-term memory (LSTM) recurrent neural network (RNN); and (iii) convolutional neural network (CNN). Performance evaluation was based on EEG data (n = 40 patients) derived from a selected set of surface EEG electrodes, which mimic the electrode layout of an implantable neurostimulation system. As for the RF input, 16 features in the time- and frequency-domains were selected. Raw EEG data were used for both CNN and RNN. Energy consumption was estimated by a platform-independent model based on the number of arithmetic operations (AOs) and memory accesses (MAs). To validate the estimated energy consumption, the RNN classifier was implemented on an ultra-low-power microcontroller.ResultsThe RNN seizure detector achieved a slightly better level of performance, with a median area under the precision-recall curve score of 0.49, compared to 0.47 for CNN and 0.46 for RF. In terms of energy consumption, RF was the most efficient algorithm, with a total of 67k AOs and 67k MAs per classification. This was followed by CNN (488k AOs and 963k MAs) and RNN (772k AOs and 978k MAs), whereby MAs contributed more to total energy consumption. Measurements derived from the hardware implementation of the RNN algorithm demonstrated a significant correlation between estimations and actual measurements.DiscussionAll three proposed seizure detection algorithms were shown to be suitable for application in implantable devices. The applied methodology for a platform-independent energy estimation was proven to be accurate by way of hardware implementation of the RNN algorithm. These findings show that seizure detection can be achieved using just a few channels with limited spatial distribution. The methodology proposed in this study can therefore be applied when designing new models for responsive neurostimulation.https://www.frontiersin.org/articles/10.3389/fneur.2021.703797/fullseizure detectionresponsive neurostimulationlow-power hardware implementationrandom forestrecurrent neural networkconvolutional neural network |
| spellingShingle | Farrokh Manzouri Marc Zöllin Simon Schillinger Simon Schillinger Matthias Dümpelmann Ralf Mikut Peter Woias Laura Maria Comella Andreas Schulze-Bonhage Andreas Schulze-Bonhage A Comparison of Energy-Efficient Seizure Detectors for Implantable Neurostimulation Devices Frontiers in Neurology seizure detection responsive neurostimulation low-power hardware implementation random forest recurrent neural network convolutional neural network |
| title | A Comparison of Energy-Efficient Seizure Detectors for Implantable Neurostimulation Devices |
| title_full | A Comparison of Energy-Efficient Seizure Detectors for Implantable Neurostimulation Devices |
| title_fullStr | A Comparison of Energy-Efficient Seizure Detectors for Implantable Neurostimulation Devices |
| title_full_unstemmed | A Comparison of Energy-Efficient Seizure Detectors for Implantable Neurostimulation Devices |
| title_short | A Comparison of Energy-Efficient Seizure Detectors for Implantable Neurostimulation Devices |
| title_sort | comparison of energy efficient seizure detectors for implantable neurostimulation devices |
| topic | seizure detection responsive neurostimulation low-power hardware implementation random forest recurrent neural network convolutional neural network |
| url | https://www.frontiersin.org/articles/10.3389/fneur.2021.703797/full |
| work_keys_str_mv | AT farrokhmanzouri acomparisonofenergyefficientseizuredetectorsforimplantableneurostimulationdevices AT marczollin acomparisonofenergyefficientseizuredetectorsforimplantableneurostimulationdevices AT simonschillinger acomparisonofenergyefficientseizuredetectorsforimplantableneurostimulationdevices AT simonschillinger acomparisonofenergyefficientseizuredetectorsforimplantableneurostimulationdevices AT matthiasdumpelmann acomparisonofenergyefficientseizuredetectorsforimplantableneurostimulationdevices AT ralfmikut acomparisonofenergyefficientseizuredetectorsforimplantableneurostimulationdevices AT peterwoias acomparisonofenergyefficientseizuredetectorsforimplantableneurostimulationdevices AT lauramariacomella acomparisonofenergyefficientseizuredetectorsforimplantableneurostimulationdevices AT andreasschulzebonhage acomparisonofenergyefficientseizuredetectorsforimplantableneurostimulationdevices AT andreasschulzebonhage acomparisonofenergyefficientseizuredetectorsforimplantableneurostimulationdevices AT farrokhmanzouri comparisonofenergyefficientseizuredetectorsforimplantableneurostimulationdevices AT marczollin comparisonofenergyefficientseizuredetectorsforimplantableneurostimulationdevices AT simonschillinger comparisonofenergyefficientseizuredetectorsforimplantableneurostimulationdevices AT simonschillinger comparisonofenergyefficientseizuredetectorsforimplantableneurostimulationdevices AT matthiasdumpelmann comparisonofenergyefficientseizuredetectorsforimplantableneurostimulationdevices AT ralfmikut comparisonofenergyefficientseizuredetectorsforimplantableneurostimulationdevices AT peterwoias comparisonofenergyefficientseizuredetectorsforimplantableneurostimulationdevices AT lauramariacomella comparisonofenergyefficientseizuredetectorsforimplantableneurostimulationdevices AT andreasschulzebonhage comparisonofenergyefficientseizuredetectorsforimplantableneurostimulationdevices AT andreasschulzebonhage comparisonofenergyefficientseizuredetectorsforimplantableneurostimulationdevices |