EEG autoregressive modeling analysis: A diagnostic tool for patients with epilepsy without epileptiform discharges
Introduction
Epilepsy is a common chronic neurological disorder affecting more than 50 million people worldwide (Carney et al., 2011). Its symptoms are transient and are caused by hypersynchronous activity in neuronal networks. Epilepsy is characterized by an enduring alteration in the brain that increases the predisposition toward seizures (Fisher et al., 2005). Distinguishing between epileptic seizures and nonepileptic events is not always straightforward. For example, syncopes and psychogenic nonepileptic seizures may imitate seizures and lead to diagnostic difficulty (Edfors and Erdal, 2008, Duncan, 2010). One study reported that the frequency of misdiagnosis of epilepsy ranges from 2% to 71%, with syncope and psychogenic nonepileptic seizures being the commonest imitators. Such misdiagnoses lead to mismanagement with antiepileptic drugs (AEDs), which can considerably affect patients’ lives (Xu et al., 2016).
Electroencephalogram (EEG) is a tool in assisting diagnosis of epilepsy. Although the appearance of epileptiform discharges (EDs) in EEG recordings is specific for epilepsy diagnosis, only 25%–56% of patients with epilepsy show EDs in their first EEG examination (Smith, 2005). To increase the positive EDs rate, a repeat EEG, sleep study, and hyperventilation stimulation are used for epilepsy diagnosis (Flink et al., 2002, Halasz et al., 2002). However, a substantial amount of patients with epilepsy still never show EDs in their EEG recordings.
Conventionally, epileptologists interpret EEG data through visual inspection. However, the interpretation of EEG signals through visual assessment is time consuming, particularly with the increased use of long-term or continuous video EEG recordings, in which hours or days of EEG data must be reviewed manually (Krumholz et al., 2007). In addition, the subjectivity of visual inspection results in varying clinical interpretations based on the EEG reader’s level of expertise (Acharya et al., 2018). The low positive rate of routine EEG studies poses another problem. A patient with epilepsy may undergo a routine EEG, and the results may be completely normal. This may be because the brains of patients with epilepsy generally do not continually emit EDs.
Recently, computer-aided systems and machine learning techniques have emerged as analytical methods in EEG (Giger, 2018). An autoregressive (AR) model learns from a series of timed steps and takes measurements from previous actions as inputs for a regression model to predict the value of the next time step (Jeon and Rabe-Hesketh, 2016). An AR model prediction error is defined as the difference between the predicted and exact values of the signal. It had been reported that in patients with epilepsy, the presence of spikes on the input signal increases the number of AR model prediction errors (Dandapat and Ray, 1997). The total amount of AR model prediction errors for each epoch is calculated and recorded to evaluate the density of spikes on the signal. The total number of AR model prediction errors for an EEG epoch increases dramatically during seizures. A comparison of AR model prediction errors with a threshold value is required to determine ictal state onset (Dandapat and Ray, 1997). According to aforementioned reports, the larger the total prediction error is, the weaker is the EEG stationarity. We hypothesized that AR prediction errors were greater in patients with epileptic seizures than nonepileptic seizures. In this study, we developed an AR model prediction error–based EEG classification method to distinguish between controls and patients with epilepsy without EDs.
Section snippets
Participants
Twenty-three patients with generalized epilepsy without EDs and 23 age-matched controls were enrolled into this study. To understand AR model prediction errors analysis more in patients with real epileptic discharges, 17 patients with absence seizure with typical 3 cycle per second spike and wave complexes EEG findings and clinical presentations were also included to compare. Epilepsy was diagnosed according to the criteria established by the International League Against Epilepsy. Every patient
Results
Twenty-three patients with generalized epilepsy without EDs were enrolled into this study. Of them, 10 were female participants. All of them experienced at least two times of seizure before EEG examinations and received repeated EEG examinations. Six of 23 patients had epileptiform discharges in their repeated EEG recordings. The presentations of generalized seizures were generalized tonic-clonic in 22 of 23 patients. The remaining one was myoclonic seizure. Regarding to antiepileptic drugs
Discussion
In this study, the AR model prediction errors were significantly higher in patients with generalized epilepsy without EDs than in controls. Similarly in another type of generalized seizure with EDs, the AR model prediction errors also revealed significantly higher in patients with absence seizure than in controls. We used an AR model prediction error–based method for the classification of patients with generalized epilepsy without EDs and controls. The developed method had an ACC, AUC, TPR, and
Conclusions
In this study, we proposed an AR model prediction error–based classification method to distinguish EEG signals between controls and patients with generalized epilepsy without EDs. The results yielded an ACC, AUC, TPR, and TNR of 85.17%, 87.54%, 89.98%, and 81.81%, respectively. An increased AR coefficient indicates nonstationary brain activity. This proposed method can help neurologists in the early diagnosis of epilepsy and may assist them in differentiating between nonepileptic paroxysmal
Acknowledgments
We wish to thank those who participated in this study. This study was supported partly by grants from Center for Smart Health Care Research of Kaohsiung Medical University (KMU-TC108B05), Kaohsiung Medical University Hospital (KMUH107-7R42), and the Ministry of Science and Technology, Taiwan (MOST 106-2314-B-037-080-MY2, MOST 107-2221-E-214-028, MOST 108-2221-E-214-020, and MOST 108-2221-E-153-010).
Declaration of Competing Interest
None of the authors have potential conflicts of interest to disclose.
References (25)
- et al.
Deep convolutional neural network for the automated detection and diagnosis of seizure using EEG signals
Comput Biol Med
(2018) Automated system for detection of epileptiform patterns in EEG by using a modified RBFN classifier
Expert Syst Appl
(2005)- et al.
Epileptic EEG detection using the linear prediction error energy
Expert Syst Appl
(2010) - et al.
Seizure prediction: methods
Epilepsy Behav
(2011) - et al.
Effect of sleep deprivation on spike-wave discharges in idiopathic generalised epilepsy: a 4 x 24 h continuous long term EEG monitoring study
Epilepsy Res
(2002) - et al.
EPILAB: a software package for studies on the prediction of epileptic seizures
J Neurosci Methods
(2011) - et al.
Frequency of a false positive diagnosis of epilepsy: A systematic review of observational studies
Seizure
(2016) Determination of autoregressive model orders for seizure detection
Turk J Electr Eng Co
(2010)- et al.
XGBoost: A scalable tree boosting system
- et al.
Spike detection in biomedical signals using midprediction filter
Med Biol Eng Comput
(1997)
Psychogenic nonepileptic seizures: diagnosis and initial management
Expert Rev Neurother
Tilt table testing in patients with suspected epilepsy
Acta Neurol Scand
Cited by (9)
Prediction of patients with idiopathic generalized epilepsy from healthy controls using machine learning from scalp EEG recordings
2023, Brain ResearchCitation Excerpt :Our XGB system showed higher values, an accuracy of 98.13% and the best AUC at 98.00%. Compared to a similar study with only 23 patients with GTCS and myoclonic seizures, XGB achieved the best values, at 85.17% and 87.54% for accuracy and AUC, respectively (Ouyang et al., 2020). Perhaps a greater number of patients and a significant variability between the different subtypes of IGE could have biased our study towards presenting higher success.
A design of bat-based optimized deep learning model for EEG signal analysis
2023, Multimedia Tools and ApplicationsLinear Random Process Model-Based EEG Classification Using Machine Learning Techniques
2023, CEUR Workshop ProceedingsA Design of Bat based Optimized Deep Learning Model for EEG Signal Analysis
2022, Research SquareVIMSNet: an effective network for visually induced motion sickness detection
2022, Signal, Image and Video Processing
- 1
Professor Chen-Sen Ouyang and Rei-Cheng Yang contribute equally as first authors.