Select for better learning: identifying high-quality training data for a multimodal cyclic transformer.

Journal: Journal of neural engineering
PMID:

Abstract

. Tonic-clonic seizures (TCSs), which present a significant risk for sudden unexpected death in epilepsy, require accurate detection to enable effective long-term monitoring. Previous studies have demonstrated the advantages of multimodal seizure detection systems in reliably detecting TCSs over extended periods. However, the effectiveness of these data-driven systems depends heavily on the availability of reliable training data.. To address this need, we propose an innovative data selection method designed to identify high-quality training samples. Our approach evaluates sample quality based on learning difficulty, classifying samples with lower learning difficulty as higher quality. We then introduce a confidence-based method to quantify the proportion of high-quality samples within the dataset.. Experimental results show that our method improves the performance of a state-of-the-art TCS detection model by 11%.. Using this data selection method, we develop a training pipeline that enhances the training process of multimodal seizure detection models.

Authors

  • Jingwei Zhang
    Department of Neurosurgery, Xiangya Hospital, Central South University, Changsha, China.
  • Zhaoyi Liu
    imec-Distrinet, Computer Science, KU Leuven, Leuven, Belgium.
  • Christos Chatzichristos
    STADIUS Center for Dynamical Systems, Signal Processing, and Data Analytics, Department of Electrical Engineering, KU Leuven, Leuven, Belgium.
  • Sam Michiels
    imec-Distrinet, Computer Science, KU Leuven, Leuven, Belgium.
  • Wim Van Paesschen
    Laboratory for Epilepsy Research, KU Leuven Biomedical Sciences Group, Leuven, Belgium.
  • Danny Hughes
    imec-Distrinet, Computer Science, KU Leuven, Leuven, Belgium.
  • Maarten De Vos
    STADIUS Center for Dynamical Systems, Signal Processing and Data Analytics-Department of Electrical Engineering (ESAT), KU Leuven, Leuven, Belgium. maarten.devos@kuleuven.be.