Neuromorphic Architecture Accelerated Automated Seizure Detection in Multi-Channel Scalp EEG

Ambati, Ravi Tej; Raja, S. Selvakumar; Alhameed, Majed; John, Titus; Arjoune, Youness; Shekhar, Raj

doi:10.3390/s22051852

Cited by 7 publications

(3 citation statements)

References 62 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Even though the data recorded using the present protocol produced reliable outcomes in signal quality, it limits the options in advanced data analytics to explore the reliability and usability of the data for clinical decision-making and monitoring. The research team is also considering a machine learning or deep learning approach to assist in seizure presence detection [ 75 ] from all the 16 channels and to categorize the severity of the identified seizure. As part of phase two of the device testing and design review process, this study also includes the device certification initiative and intellectual property (IP) filling to ensure OptiEEG would be clinically applicable.…”

Section: Discussionmentioning

confidence: 99%

Assessment of a 16-Channel Ambulatory Dry Electrode EEG for Remote Monitoring

Shivaraja

Remli

Kamal

et al. 2023

Sensors

View full text Add to dashboard Cite

Ambulatory EEGs began emerging in the healthcare industry over the years, setting a new norm for long-term monitoring services. The present devices in the market are neither meant for remote monitoring due to their technical complexity nor for meeting clinical setting needs in epilepsy patient monitoring. In this paper, we propose an ambulatory EEG device, OptiEEG, that has low setup complexity, for the remote EEG monitoring of epilepsy patients. OptiEEG’s signal quality was compared with a gold standard clinical device, Natus. The experiment between OptiEEG and Natus included three different tests: eye open/close (EOC); hyperventilation (HV); and photic stimulation (PS). Statistical and wavelet analysis of retrieved data were presented when evaluating the performance of OptiEEG. The SNR and PSNR of OptiEEG were slightly lower than Natus, but within an acceptable bound. The standard deviations of MSE for both devices were almost in a similar range for the three tests. The frequency band energy analysis is consistent between the two devices. A rhythmic slowdown of theta and delta was observed in HV, whereas photic driving was observed during PS in both devices. The results validated the performance of OptiEEG as an acceptable EEG device for remote monitoring away from clinical environments.

show abstract

Section: Discussionmentioning

confidence: 99%

Assessment of a 16-Channel Ambulatory Dry Electrode EEG for Remote Monitoring

Shivaraja

Remli

Kamal

et al. 2023

Sensors

View full text Add to dashboard Cite

show abstract

“…Using the Bonn EEG dataset, they obtained an accuracy of 93.9% from CNNs and 97.2% from bidirectional LSTM. However, using RCE networks DWT features, another study [10] obtained a specificity of 98.77%.…”

Section: Traditional Approachesmentioning

confidence: 95%

CAD system for epileptic seizure detection from EEG through image processing and SURF-BOF technique

Alshayeji

2023

Mach. Learn.: Sci. Technol.

View full text Add to dashboard Cite

Epilepsy is one of the most debilitating neurological diseases that abruptly alters a person's way of life. Manual diagnosis is a laborious and time-consuming task prone to human error; therefore, automating this task by developing an intelligent system is necessary. Existing deep learning (DL) models require high training time, large datasets, and machines with more memory and processing power. In addition, owing to the black-box nature of DL models, no one can determine the features that the network prefers for classification decisions. To overcome these challenges, this paper proposes an accurate, automatic, and fast intelligent system for epilepsy detection using a computer-aided diagnosis (CAD)-2D machine learning (ML) framework. Existing ML models struggle to produce reliable and acceptable diagnostic results owing to the low amplitude and nonstationary nature of electroencephalograms (EEG), particularly in clinical situations where environmental influences are almost impossible to eliminate. The proposed model was built on the CHB-MIT dataset, and it represents the first study that employs the speeded-up robust feature (SURF) bag of features (BOF) technique for this application, which generates local features from the spectrogram images of the respective 1D EEG signal inputs. In addition, DL features were extracted from the spectrogram images for model performance comparisons. Both features were used separately to train the ML classifiers. Implementing SURF offers fast computation and makes the model invariant to distortions, noise, scaling, and so on. Therefore, the proposed model is more suitable for real-time applications, proposed ML framework provides an enhanced accuracy of 99.78% from the SVM-RBF classifier, along with 99.56% sensitivity, 100% specificity, and an error rate of 0.22%. The higher detection accuracy demonstrated the effectiveness of the proposed framework for medical disease diagnosis applications.

show abstract

“…The source reconstruction technique enables us to differentiate between generalized and focal seizures. The latter can be treated with brain surgery [90][91][92]. The source reconstruction technique can localize an epileptogetic focus and guide the neurointervention.…”

Section: Brain Electrophysiologic Activity In Seizure Differentiation...mentioning

confidence: 99%

Automatic Detection and Classification of Epileptic Seizures from EEG Data: Finding Optimal Acquisition Settings and Testing Interpretable Machine Learning Approach

Statsenko,

Babushkin,

Talako

et al. 2023

Biomedicines

View full text Add to dashboard Cite

Deep learning (DL) is emerging as a successful technique for automatic detection and differentiation of spontaneous seizures that may otherwise be missed or misclassified. Herein, we propose a system architecture based on top-performing DL models for binary and multigroup classifications with the non-overlapping window technique, which we tested on the TUSZ dataset. The system accurately detects seizure episodes (87.7% Sn, 91.16% Sp) and carefully distinguishes eight seizure types (95–100% Acc). An increase in EEG sampling rate from 50 to 250 Hz boosted model performance: the precision of seizure detection rose by 5%, and seizure differentiation by 7%. A low sampling rate is a reasonable solution for training reliable models with EEG data. Decreasing the number of EEG electrodes from 21 to 8 did not affect seizure detection but worsened seizure differentiation significantly: 98.24 ± 0.17 vs. 85.14 ± 3.14% recall. In detecting epileptic episodes, all electrodes provided equally informative input, but in seizure differentiation, their informative value varied. We improved model explainability with interpretable ML. Activation maximization highlighted the presence of EEG patterns specific to eight seizure types. Cortical projection of epileptic sources depicted differences between generalized and focal seizures. Interpretable ML techniques confirmed that our system recognizes biologically meaningful features as indicators of epileptic activity in EEG.

show abstract

Neuromorphic Architecture Accelerated Automated Seizure Detection in Multi-Channel Scalp EEG

Cited by 7 publications

References 62 publications

Assessment of a 16-Channel Ambulatory Dry Electrode EEG for Remote Monitoring

Assessment of a 16-Channel Ambulatory Dry Electrode EEG for Remote Monitoring

CAD system for epileptic seizure detection from EEG through image processing and SURF-BOF technique

Automatic Detection and Classification of Epileptic Seizures from EEG Data: Finding Optimal Acquisition Settings and Testing Interpretable Machine Learning Approach

Contact Info

Product

Resources

About