Cross-Subject Tinnitus Diagnosis Based on Multi-Band EEG Contrastive Representation Learning

Wang, Chang‐Dong; Zhu, Xi-Ran; Zhou, Xueqing; Li, Jiahong; Lan, Liping; Huang, Dong; Zheng, Yuxin; Cai, Yixi

doi:10.1109/jbhi.2023.3264521

Cited by 3 publications

(1 citation statement)

References 52 publications

(52 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Contrastive learning is an SSL method that learns general feature representation of samples by comparing the data with positive and negative samples in the feature space and has demonstrated remarkable accomplishments in computer vision, natural language processing, and other fields [37][38][39][40]. However, few studies have applied contrastive learning techniques to EEG decoding [41][42][43]. In [44], an emotion recognition method ECNN-C was proposed, which used an innovative convolution module and introduced contrastive learning into CNN, developing a hybrid loss model with contrastive loss and cross-entropy loss.…”

Section: Introductionmentioning

confidence: 99%

Self-supervised contrastive learning for EEG-based cross-subject motor imagery recognition

Li,

Sun

et al. 2024

J. Neural Eng.

View full text Add to dashboard Cite

Objective. The extensive application of electroencephalography (EEG) in brain-computer interfaces (BCIs) can be attributed to its non-invasive nature and capability to offer high-resolution data. The acquisition of EEG signals is a straightforward process, but the datasets associated with these signals frequently exhibit data scarcity and require substantial resources for proper labeling. Furthermore, there is a significant limitation in the generalization performance of EEG models due to the substantial inter-individual variability observed in EEG signals. Approach. To address these issues, we propose a novel self-supervised contrastive learning framework for decoding motor imagery (MI) signals in cross-subject scenarios. Specifically, we design an encoder combining convolutional neural network and attention mechanism. In the contrastive learning training stage, the network undergoes training with the pretext task of data augmentation to minimize the distance between pairs of homologous transformations while simultaneously maximizing the distance between pairs of heterologous transformations. It enhances the amount of data utilized for training and improves the network’s ability to extract deep features from original signals without relying on the true labels of the data. Main results. To evaluate our framework’s efficacy, we conduct extensive experiments on three public MI datasets: BCI IV IIa, BCI IV IIb, and HGD datasets. The proposed method achieves cross-subject classification accuracies of 67.32%, 82.34%, and 81.13% on the three datasets, demonstrating superior performance compared to existing methods. Significance. Therefore, this method has great promise for improving the performance of cross-subject transfer learning in MI-based BCI systems.

show abstract

Section: Introductionmentioning

confidence: 99%

Self-supervised contrastive learning for EEG-based cross-subject motor imagery recognition

Li,

Sun

et al. 2024

J. Neural Eng.

View full text Add to dashboard Cite

show abstract

Manifold Learning-Based Common Spatial Pattern for EEG Signal Classification

Cai,

Zhang,

Yang

et al. 2024

IEEE J. Biomed. Health Inform.

View full text Add to dashboard Cite

Personalized Sound Therapy Combined with Low and High-Frequency Electromagnetic Stimulation for Chronic Tinnitus

Francavilla,

Marzocchella,

Alagna

et al. 2024

JPM

View full text Add to dashboard Cite

This study investigates a novel multimodal treatment for chronic tinnitus, a condition that significantly affects quality of life, by combining personalized sound therapy with both low- and high-frequency electromagnetic wave stimulation. Conducted at Tor Vergata University Hospital in Rome, the research involved 55 patients and employed a portable medical device for therapy delivery. Treatment effectiveness was measured through the Tinnitus Functional Index (TFI), Tinnitus Handicap Inventory (THI), Visual Analogue Scale (VAS), Hyperacusis Questionnaire (HQ), and Short Form-36 Health Survey (SF-36), encompassing initial sound therapy and subsequent multimodal treatment phases. Remarkably, 73% of participants experienced notable improvements in TFI scores, with 39% reporting a significant enhancement of 13 points or more. This improvement was mirrored in secondary outcomes like THI, VAS, and HQ scores, along with certain SF-36 domains, indicating enhanced life quality and reduced tinnitus distress. The study underscored high compliance and no adverse effects, suggesting the combined therapy’s promising potential in chronic tinnitus management. The findings advocate for further research to discern the distinct contributions of each treatment modality, positing that this innovative approach could ameliorate tinnitus symptoms and improve patient well-being, confirming its safety and efficacy.

show abstract

Cross-Subject Tinnitus Diagnosis Based on Multi-Band EEG Contrastive Representation Learning

Cited by 3 publications

References 52 publications

Self-supervised contrastive learning for EEG-based cross-subject motor imagery recognition

Self-supervised contrastive learning for EEG-based cross-subject motor imagery recognition

Manifold Learning-Based Common Spatial Pattern for EEG Signal Classification

Personalized Sound Therapy Combined with Low and High-Frequency Electromagnetic Stimulation for Chronic Tinnitus

Contact Info

Product

Resources

About