HFO Detection in Epilepsy: A Stacked Denoising Autoencoder and Sample Weight Adjusting Factors-Based Method

Abstract

High-frequency oscillations (HFOs) recorded by the intracranial electroencephalography (iEEG) are the promising biomarkers of epileptogenic zones. Accurate detection of HFOs is the key to pre-operative assessment for epilepsy. Due to the subjective bias caused by manual features and the class imbalance between HFOs and false HFOs, it is difficult to obtain satisfactory detection performance by the existing methods. To solve these problems, we put forward a novel method to accurately detect HFOs based on the stacked denoising autoencoder (SDAE) and the ensemble classifier with sample weight adjusting factors. First, the adjustable threshold of Hilbert envelopes is proposed to isolate the events of interest (EoIs) from background activities. Then, the SDAE network is utilized to automatically extract features of EoIs in the time-frequency domain. Finally, the AdaBoost-based support vector machine ensemble classifier with sample weight adjusting factors is devised to separate HFOs from EoIs by using the extracted features. These adjusting factors are used to solve the class imbalance problem by adjusting sample weights when learning the base classifiers. Our HFO detection method is evaluated by using clinical iEEG data recorded from 20 patients with medically refractory epilepsy. The experimental results show that our detection method outperforms some existing methods in terms of sensitivity and false discovery rate. In addition, the HFOs detected by our method are effective for localizing seizure onset zones.

An approach for reliably identifying high‐frequency oscillations and reducing false‐positive detections

“…According to the evaluation of HFO detection, our algorithm reached the accuracy rate of 92% and a precision rate of 99%. This result is close to the machine learning approaches with powerful recognition capabilities brought by neural networks 45–50 . However, the machine learning models hold an end‐to‐end design, which lack the interpretation and mathematical description of HFOs important to clinical utilization 45–48 .…”

[Retracted] Automatic Detection of High‐Frequency Oscillations Based on an End‐to‐End Bi‐Branch Neural Network and Clinical Cross‐Validation

“…In terms of the cross-subject validation, in 2019, Zuo et al [ 28 ] proposed to convert the collected candidate HFOs into a two-dimensional gray-scale matrix and then use a stacked CNN to further distinguish the candidate events, which achieved a sensitivity of 77.04% in ripples and 83.23% in fast ripples and a specificity of 72.27% in ripples and 79.36% in fast ripples. In 2021, Wu et al [ 29 ] proposed a novel detector based on the stacked denoising autoencoder (SDAE) and the ensemble classifier with sample weight adjusting factors, which achieved a sensitivity of 92.4% in ripples and 90.3% in fast ripples and an FDR of 9.2% in ripples and 10.7% in fast ripples. Besides, in 2021, Wang et al [ 47 ] proposed an algorithm to calculate the dynamic baseline based on the maximum distributed peak points to automatically detect HFOs and achieved a sensitivity of 82.666% and a specificity of 63.352%.…”

“…However, it is difficult for these methods to distinguish HFOs from some artifacts, such as spikes, pulse-like artifacts, and signals with harmonics [ 20 ]. To alleviate this problem, two-stage methods are proposed to further explore the signal characteristics by adding a supervised classifier or an unsupervised clustering after an initial detector [ 25 – 29 ]. Specifically, real HFOs can be identified from the candidate events isolated from background activities by a stacked CNN [ 28 ] or a stacked denoising autoencoder [ 29 ], while in terms of time-frequency diagram, the two-stage automatic detection paradigm has always been used.…”

“…To alleviate this problem, two-stage methods are proposed to further explore the signal characteristics by adding a supervised classifier or an unsupervised clustering after an initial detector [ 25 – 29 ]. Specifically, real HFOs can be identified from the candidate events isolated from background activities by a stacked CNN [ 28 ] or a stacked denoising autoencoder [ 29 ], while in terms of time-frequency diagram, the two-stage automatic detection paradigm has always been used. In particular, it is the simple 2d-CNN structure that is often used as a time-frequency image feature extractor after an initial detector [ 25 , 30 – 32 ].…”

Implementation of a Morphological Filter for Removing Spikes from the Epileptic Brain Signals to Improve Identification Ripples

“…Various HFO detection algorithms have already been proposed in the most current literature (see inter alia: [ 84 , 85 , 86 , 87 , 88 , 89 ]). These simple algorithms used as the first step band-pass filtering and some statistical measurements, such as, among others, RMS (root mean square) [ 71 ], line length [ 87 , 90 ] or Hilbert transform [ 91 ].…”