Seizure Detection: A Low Computational Effective Approach without Classification Methods

Abstract

Epilepsy is a severe neurological disorder that is usually diagnosed by using an electroencephalogram (EEG). However, EEG signals are complex, nonlinear, and dynamic, thus generating large amounts of data polluted by many artefacts, lowering the signal-to-noise ratio, and hampering expert interpretation. The traditional seizure-detection method of professional review of long-term EEG signals is an expensive, time-consuming, and challenging task. To reduce the complexity and cost of the task, researchers have developed several seizure-detection approaches, primarily focusing on classification systems and spectral feature extraction. While these methods can achieve high/optimal performances, the system may require retraining and following up with the feature extraction for each new patient, thus making it impractical for real-world applications. Herein, we present a straightforward manual/automated detection system based on the simple seizure feature amplification analysis to minimize these practical difficulties. Our algorithm (a simplified version is available as additional material), borrowing from the telecommunication discipline, treats the seizure as the carrier of information and tunes filters to this specific bandwidth, yielding a viable, computationally inexpensive solution. Manual tests gave 93% sensitivity and 96% specificity at a false detection rate of 0.04/h. Automated analyses showed 88% and 97% sensitivity and specificity, respectively. Moreover, our proposed method can accurately detect seizure locations within the brain. In summary, the proposed method has excellent potential, does not require training on new patient data, and can aid in the localization of seizure focus/origin.

Keywords: EEG; feature identification; low computation method; seizure detection.

Figure 1

Common stages of automatic seizure-detection process based on feature extraction and classification.

Figure 2

Logical flowchart of the proposed seizure-detection algorithm. The loop is repeated once for every dataset of one hour in length. An expert can review the results of both manual and automated data to identify any variations in false alarms and make corrections within a few minutes.

Figure 3

Direct comparison of raw EEG (blue trace) and preprocessed EEG signal. (a) Direct comparison of raw EEG (blue trace) and pre-processed EEG signal during artefact. (b) Direct comparison of raw EEG (blue trace) and pre-processed EEG signal artefact clean section.

Figure 4

Output for chb01_01, using the proposed seizure-detection algorithm, shows no seizure activity.

Figure 5

Feature extraction showing the onset of seizure for chb01_03, using the proposed seizure-detection algorithm. The highlighted red box shows the region of seizure.

Figure 6

Feature showing the onset of seizure marked in the output window for chb01_26, using the proposed seizure-detection algorithm.

Figure 7

Location identification for chb01_03 using the proposed algorithm. Channels 1, 5, 9, and 14 show the amplitude level above the set threshold corresponding to the frontal-lobe seizure.

Figure 8

(a) Test results for the whole CHB MIT EEG dataset. (b) Test results for the whole CHB MIT EEG dataset (F_M is F-measure; G_M is geometric mean). (c) Automated test results for the whole CHB MIT EEG dataset.

Figure 8

(a) Test results for the whole CHB MIT EEG dataset. (b) Test results for the whole CHB MIT EEG dataset (F_M is F-measure; G_M is geometric mean). (c) Automated test results for the whole CHB MIT EEG dataset.

Figure 9

Time-based performance of the proposed algorithm: (TPR is total positive rate) and (FP/hour is false positives per hour. Note: The figure shows the total positive and false-positive rates per hour for all the individual trials of patient data. In the dataset, a single patient may have a varied number of 15 to 42 data files. Each data file is named according to the number of trials; that is, Patient 1 has 42 data files or trials.