Abstract
Deep learning techniques have led to significant advancements in seizure prediction research. However, corresponding used benchmarks are not uniform in published results. Moreover, inappropriate training and evaluation processes used in various work create overfitted models, making prediction performance fluctuate or unreliable. In this study, we analyzed the various data preparation methods, dataset partition methods in related works, and explained the corresponding impacts to the prediction algorithms. Then we applied a robust processing procedure that considers the appropriate sampling parameters and the leave-one-out cross-validation method to avoid possible overfitting and provide prerequisites for ease benchmarking. Moreover, a deep learning architecture takes advantage of a one-dimension convolutional neural network and a bi-directional long short-term memory network is proposed for seizure prediction. The architecture achieves 77.6% accuracy, 82.7% sensitivity, and 72.4% specificity, and it outperforms the indicators of other prior-art works. The proposed model is also hardware friendly; it has 6.274 k parameters and requires only 12.825 M floating-point operations, which is advantageous for memory and power constrained device implementations.
Highlights
Time are two important parameters for sampling
We propose in this work a processing procedure as a reference for training reliably seizure prediction algorithms and facilitating fair benchmarking of successive prediction methods
A one-dimensional CNN with Bi-LSTM network is proposed and evaluated on CHB-MIT database shows that it could help improve prediction performance, and the demonstrated power and cost efficiency are advantageous for implant devices
Summary
Time are two important parameters for sampling. As shown in Fig. 1b, the SOP is the interval where the seizure is expected to occur, the period between the alarm and the beginning of the SOP is the S PH16. The definition of leading seizure has been entirely overlooked in some studies[8, 10, 18], and even in those studies that consider this concept, the value of seizure-free time (T) varied from 30 min to 4 h 16,17,19. The number of available seizures are not equal due to different definitions of leading seizures, which result in a different number of cross-validation experiments. These validation methods are so different that the reliability of the model cannot be guaranteed. Several issues are discussed based on the experimental results in "Discussion" section, which greatly influence model performance, including the different choice of sampling parameters and dataset partition methods.
Sign-in/Register to view highlights & summary Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
