A hierarchical approach for online temporal lobe seizure detection in long-term intracranial EEG recordings

Liang, Sheng; Chen, Yi Chun; Wang, Yu Lin; Chen, Pin Tzu; Yang, Chia-Hsiang; Chiueh, Herming

doi:10.1088/1741-2560/10/4/045004

Cited by 30 publications

(9 citation statements)

References 38 publications

(52 reference statements)

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…After dynamic system theory was used to develop the framework of viewing epilepsy as a “dynamical disease,” the seizure prediction/detection/anticipation and control has been investigated with unparalleled enthusiasm. Over the Past decade, many seizure prediction/detection algorithms have been proposed to anticipate seizures . However, these algorithms were tested offline using high performance computers and the detection sensitivities ranged from 62% to 98%.…”

Section: Discussionmentioning

confidence: 99%

Rapid brief feedback intracerebral stimulation based on real‐time desynchronization detection preceding seizures stops the generation of convulsive paroxysms

et al. 2015

View full text Add to dashboard Cite

SUMMARYObjective: To investigate the abortion of seizure generation using "minimal" intervention in hippocampi using two rat models of human temporal lobe epilepsy. Methods: The recording or stimulation electrodes were implanted into both hippocampi (CA1 area). Using the kainic acid (chronic: experiment duration 24 days) and the 4-aminopyridine (acute: experiment duration 2 h) models of paroxysms in rats, a real-time feedback stimulation paradigm was implemented, which triggered a short periodic electrical stimulus (5 Hz for 5 s) upon detecting a seizure precursor. Our seizure precursor detection algorithm relied on the monitoring of the real-time phase synchronization analysis, and detected/anticipated electrographic seizures as early as a few seconds to a few minutes before the behavioral and electrographic seizure onset, with a very low false-positive rate of the detection. Results: The baseline mean seizure frequencies were 5.39 seizures per day (chronic) and 13.2 seizures per hour (acute). The phase synchrony analysis detected 88% (434 of 494) of seizures with a mean false alarm of 0.67 per day (chronic) and 83% (86 of 104) of seizures with a mean false alarm of 0.47 per hour (acute). The feedback stimulation reduced the seizure frequencies to 0.41 seizures per day (chronic) and 2.4 seizures per hour (acute). Overall, the feedback stimulation paradigm reduced seizure frequency by a minimum of 80% to a maximum of 100% in 10 rats, with 83% of the animals rendered seizure-free. Significance: This approach represents a simple and efficient manner for stopping seizure development. Because of the short on-demand stimuli, few or no associated side effects are expected in clinical application in patients with epilepsy. Abnormal synchrony patterns are common features in epilepsy and other neurologic and psychiatric syndromes; therefore, this type of feedback stimulation paradigm could be a novel therapeutic modality for use in various neurologic and psychiatric disorders.

show abstract

Section: Discussionmentioning

confidence: 99%

Rapid brief feedback intracerebral stimulation based on real‐time desynchronization detection preceding seizures stops the generation of convulsive paroxysms

et al. 2015

View full text Add to dashboard Cite

show abstract

“…• Hjorth Descriptors, calculate the Hjorth parameters for activity, mobility, and complexity (Liang et al, 2013) (Hjorth, 1970).…”

Section: Signals Extractions and Processingmentioning

confidence: 99%

The Role of Personalization and Multiple EEG and Sound Features Selection in Real Time Sonification for Neurofeedback

Mealla¹,

Oliveira²,

Marimon³

et al. 2014

Proceedings of the International Conference on Physiological Computing Systems

View full text Add to dashboard Cite

Abstract:The field of physiology-based interaction and monitoring is developing at a fast pace. Emerging applications like fatigue monitoring often use sound to convey complex dynamics of biological signals and to provide an alternative, non-visual information channel. Most Physiology-to-Sound mappings in such auditory displays do not allow customization by the end-users. We designed a new sonification system that can be used for extracting, processing and displaying Electroencephalography data (EEG) with different sonification strategies. The system was validated with four user groups performing alpha/theta neurofeedback training (a/t) for relaxation that varied in feedback personalization (Personalized/Fixed) and a number of sonified EEG features (Single/Multiple). The groups with personalized feedback performed significantly better in their training than fixed mappings groups, as shown by both subjective ratings and physiological indices. Additionally, the higher number of sonified EEG features resulted in deeper relaxation than when training with single feature feedback. Our results demonstrate the importance of adaptation and personaliziation of EEG sonification according to particular applications, in our case, to a/t neurofeedback. Our experimental approach shows how user performance can be used for validating different sonification strategies.

show abstract

“…It has considerable economic implications in terms of health-care needs, and sometimes, may result in death of matured persons. Around 1% of the world's population is affected by epilepsy, and nearly 25% of them cannot be treated effectively by available therapies due to resistance to drugs [1]. However, the resection of the epileptogenic zone is rather difficult due to the overlay with other eloquent areas responsible for language, primary motor and vision.…”

Section: Introductionmentioning

confidence: 99%

Discrimination of scalp EEG signals in wavelet transform domain and channel selection for the patient-invariant seizure detection

Das

Pantho

Bhuiyan

2015

2015 International Conference on Electrical &Amp; Electronic Engineering (ICEEE)

View full text Add to dashboard Cite

In this paper, a statistical method of classifying Electroencephalogram(EEG) data for automatic detection of epileptic seizure is carried out using a publicly available scalp EEG database. The classification is carried out to distinguish the seizure segments from the non-seizure ones. The higher order moments (specifically variance) have been calculated in various sub-bands in the wavelet transform domain and utilized as the discriminating feature in the Support Vector Machine(SVM) classifier. The method is tested on 175 hours of continuous EEG data from five patients and on an average, 99% accuracy has been achieved with very high values of sensitivity and specificity. Furthermore, on the basis of the figure of merits, for their excellent performance for all the patients, seven channels have been selected for the patient-invariant seizure detection which might help the electroencephalographers reducing their laborious job of monitoring the EEG data from all the channels.

show abstract

A hierarchical approach for online temporal lobe seizure detection in long-term intracranial EEG recordings

Cited by 30 publications

References 38 publications

Rapid brief feedback intracerebral stimulation based on real‐time desynchronization detection preceding seizures stops the generation of convulsive paroxysms

Rapid brief feedback intracerebral stimulation based on real‐time desynchronization detection preceding seizures stops the generation of convulsive paroxysms

The Role of Personalization and Multiple EEG and Sound Features Selection in Real Time Sonification for Neurofeedback

Discrimination of scalp EEG signals in wavelet transform domain and channel selection for the patient-invariant seizure detection

Contact Info

Product

Resources

About