What is the clinical evidence from this study?

Study design: Other. Population: Healthy subjects (SSVEP-BCI) (n=105). Intervention: eTRCA + sbCNN framework vs. eTRCA, sbCNN, and other state-of-the-art models. Primary outcome: Classification accuracy at 1.0s data length (Benchmark dataset) (p=<0.05).

January 2, 2025Open Access

Enhancing the performance of SSVEP-based BCIs by combining task-related component analysis and deep neural network

Q: What are the key findings of this study?

The eTRCA + sbCNN framework significantly improved SSVEP classification performance compared to individual models, achieving 96.19% accuracy on the Benchmark dataset at 1.0s data length.

Q: What does this research mean for the field?

Combining an ensemble task-related component analysis (eTRCA) algorithm with a sub-band convolutional neural network (sbCNN) significantly improves SSVEP classification performance compared to individual models, achieving 96.19% accuracy on the Benchmark dataset at 1.0s data length. Novelty: ClaimNovelty.METHODOLOGICAL. Consensus alignment: ConsensusAlignment.NEUTRAL.

QWQingguo WeiNorth University of China CLChang LiHunan University YWYijun WangShanghai Jiao Tong University

Key Result

The eTRCA + sbCNN framework significantly improved SSVEP classification performance compared to individual models, achieving 96.19% accuracy on the Benchmark dataset at 1.0s data length.

Structured PICO

Population

105 healthy subjects from two publicly available SSVEP datasets (Benchmark: n=35, 17 females, average age 22 years; BETA: n=70, 42 males, average age 25 years)

Intervention

eTRCA + sbCNN classification framework for recognizing the frequency of SSVEP signals

Comparator

eTRCA alone, sbCNN alone, and other state-of-the-art models

Outcome

Classification accuracy and Information Transfer Rate (ITR)

Combining ensemble task-related component analysis and a sub-band convolutional neural network significantly improves the classification accuracy and information transfer rate of SSVEP-based brain-computer interfaces.

Main Result

Absolute Event Rate: 96.19% vs 88.43%

p-value: p=<0.05

Limitations

The training time is too long when the sbCNN model is trained using transfer learning (about 2.4 hours).
Only an offline analysis of the proposed algorithm was conducted in this study.

Abstract

Steady-State Visually Evoked Potential (SSVEP) signals can be decoded by either a traditional machine learning algorithm or a deep learning network. Combining the two methods is expected to enhance the performance of an SSVEP-based brain-computer interface (BCI) by exploiting their advantages. However, an efficient strategy for integrating the two methods has not yet been established. To address this issue, we propose a classification framework named eTRCA + sbCNN that combines an ensemble task-related component analysis (eTRCA) algorithm and a sub-band convolutional neural network (sbCNN) for recognizing the frequency of SSVEP signals. The two models are first trained separately, then their classification score vectors are added together, and finally the frequency corresponding to the maximal summed score is decided as the frequency of SSVEP signals. The proposed framework can effectively exploit the complementarity between the two kinds of feature signals and significantly improve the classification performance of SSVEP-based BCIs. The performance of the proposed method is validated on two SSVEP BCI datasets and compared with that of eTRCA, sbCNN and other state-of-the-art models. Experimental results indicate that the proposed method significantly outperform the compared algorithms, and thus helps to promote the practical application of SSVEP- BCI systems.

KI fragen

Bookmark

View Full Paper