EUVIMEDEUVIMED entdecken
Crossref

Discrete Wavelet Convolution for Learnable Time-Frequency Representation with Application to Seizure Prediction

Weisen Lu iD, Haotian Li iD, Guoyang Liu iD, Yi Li iD, Weidong Zhou iD

International Journal of Neural Systems · 30.12.2026 · 36 (13)

DOI: 10.1142/s0129065726500498 ↗

Kurz gefasst

Fachartikel in der Zeitschrift International Journal of Neural Systems, online erschienen am 24.06.2026. Verfasst von Weisen Lu und 4 weiteren Autor:innen. Beteiligte Einrichtung: School of Integrated Circuits, Shandong University, Jinan 250100, P. R. China. Hauptthemen laut MEDLINE: Electroencephalography, Wavelet Analysis, Seizures, Deep Learning. Die Arbeit nennt 62 Literaturangaben. Gefördert von: the National Natural Science Foundation of China, the Key Program of the Natural Science Foundation of Shandong Province.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Accurate and adaptive time-frequency representation is essential for analyzing nonstationary signals in critical applications, such as epileptic seizure prediction utilizing electroencephalogram (EEG) data. However, existing deep learning approaches often suffer from two major limitations: the reliance on fixed, nonadaptive feature extraction methods and the lack of model interpretability. To bridge these gaps, we propose Adaptive Discrete Wavelet Convolution (ADWConv), a general-purpose trainable multi-level wavelet decomposition front-end, with validation in an end-to-end seizure prediction framework. ADWConv parameterizes the low-frequency and high-frequency filters within each decomposition level and decouples their kernel shapes from level-specific scaling factors, enabling task-related adaptive weighting of frequency bands. This end-to-end learning is guided by a joint optimization strategy, which incorporates a regularization loss with wavelet priors to ensure the learned filters are both discriminative and physically interpretable. Extensive experiments were conducted on the public CHB-MIT database and the private SH-SDU database. On the CHB-MIT database, the ADWConv-based model attained an event-based sensitivity of 98.65%, while maintaining a false prediction rate (FPR) of only 0.014/h. Comparable results were also demonstrated on the SH-SDU database, attaining 96.53% event-based sensitivity with 0.028/h FPR. Ablation and visualization analyses confirm that learning the wavelet kernels is the primary driver of performance gains and supports task-specific adaptation of discriminative frequency bands. Overall, ADWConv improves the feature-level interpretability of a deep learning framework for seizure prediction while maintaining competitive accuracy and robustness.

Schlagwörter

MeSH-Schlagwörter (MEDLINE)

ElectroencephalographyWavelet AnalysisSeizuresDeep LearningSignal Processing, Computer-AssistedHumansPrediction Algorithms

Stichwörter der Autor:innen

Deep learningdiscrete wavelet transform (DWT)electroencephalogram (EEG)interpretabilityseizure prediction

Autor:innen und Institutionen

  • School of Integrated Circuits, Shandong University, Jinan 250100, P. R. China.
  • Shandong Key Laboratory of Next-Generation Semiconductor Technology and Systems, Shandong University, Jinan 250100, P. R. China.
  • Weill Cornell Medicine, Cornell University, New York 10065, State of New York, United States.

ORCID-Profile: Weisen Lu, Haotian Li, Guoyang Liu, Yi Li, Weidong Zhou

Förderung

  • the National Natural Science Foundation of China
  • the Key Program of the Natural Science Foundation of Shandong Province

Zitieren

APA

Lu, W., Li, H., Liu, G., Li, Y., & Zhou, W. (2026). Discrete Wavelet Convolution for Learnable Time-Frequency Representation with Application to Seizure Prediction. International Journal of Neural Systems, 36(13). https://doi.org/10.1142/s0129065726500498

Vancouver

Lu W, Li H, Liu G, Li Y, Zhou W. Discrete Wavelet Convolution for Learnable Time-Frequency Representation with Application to Seizure Prediction. International Journal of Neural Systems. 2026;36(13). doi:10.1142/s0129065726500498

BibTeX
@article{Lu2026,
  author = {Lu, Weisen and Li, Haotian and Liu, Guoyang and Li, Yi and Zhou, Weidong},
  title = {Discrete Wavelet Convolution for Learnable Time-Frequency Representation with Application to Seizure Prediction},
  journal = {International Journal of Neural Systems},
  year = {2026},
  volume = {36},
  number = {13},
  doi = {10.1142/s0129065726500498}
}
Kostenlos starten →