Lincong Pan

Lincong Pan

Auto-generated profile, based on publicly available information

Lincong Pan

Tianjin University, Academy of Medical Engineering and Translational Medicine, Tianjin University, School of Precision Instruments and Optoelectronics Engineering, Tianjin University, Tianjin, 300072, CHINA

Research Focus & Metrics

Academic Career Length

4Years (2023-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
22 November 2023 in Journal of Neural Engineering
Verified
Journal Article
Closed access
PDF file
LPLincong Pan
Lincong Pan
KWKun Wang
Kun Wang
LXLichao Xu
Lichao Xu
XSXinwei Sun
Xinwei Sun
WYWeibo Yi
Weibo Yi
DMDong Ming
Dong Ming
Objective. Brain–computer interfaces (BCIs) enable a direct communication pathway between the human brain and external devices, without relying on the traditional peripheral nervous and musculoskeletal systems. Motor imagery (MI)-based BCIs have attracted significant interest for their potential in motor rehabilitation. However, current algorithms fail to account for the cross-session variability of electroencephalography signals, limiting their practical application. Approach. We proposed a Riemannian geometry-based adaptive boosting and voting ensemble (RAVE) algorithm to address this issue. Our approach segmented the MI period into multiple sub-datasets using a sliding window approach and extracted features from each sub-dataset using Riemannian geometry. We then trained adaptive boosting (AdaBoost) ensemble learning classifiers for each sub-dataset, with the final BCI output determined by majority voting of all classifiers. We tested our proposed RAVE algorithm and eight other competing algorithms on four datasets (Pan2023, BNCI001-2014, BNCI001-2015, BNCI004-2015). Main results. Our results showed that, in the cross-session scenario, the RAVE algorithm outperformed the eight other competing algorithms significantly under different within-session training sample sizes. Compared to traditional algorithms that involved a large number of training samples, the RAVE algorithm achieved similar or even better classification performance on the datasets (Pan2023, BNCI001-2014, BNCI001-2015), even when it did not use or only used a small number of within-session training samples. Significance. These findings indicate that our cross-session decoding strategy could enable MI-BCI applications that require no or minimal training process.

22 April 2025 in Neural Networks
Verified
Journal Article
Closed access
LPLincong Pan
Lincong Pan
KWKun Wang
Kun Wang
YHYongzhi Huang
Yongzhi Huang
XSXinwei Sun
Xinwei Sun
JMJiayuan Meng
Jiayuan Meng
DMDong Ming
Dong Ming

10 April 2025 in TechRxiv
Verified
Preprint
Open access
LPLincong Pan
Lincong Pan
KWKun Wang
Kun Wang
WYWeibo Yi
Weibo Yi
YZYang Zhang
Yang Zhang
MXMinpeng Xu
Minpeng Xu
DMDong Ming
Dong Ming
Motor imagery brain-computer interfaces (MI-BCIs) are crucial for neuro-rehabilitation and assistive therapy, but decoding motor imagery electroencephalography (EEG) signals is challenged by non-stationarity, low signal-to-noise ratio, and cross-session variability. Methods: We propose the common temporal-spectral-spatial patterns (CTSSP) algorithm to address these challenges. The method introduces: (1) a multi-time-window design to capture dynamic neural patterns, (2) an adaptive finite impulse response (FIR) filter bank to suppress noise and amplify task-relevant features across time windows and channels, and (3) a regularization constraint to improve robustness and prevent overfitting. Results: CTSSP outperforms existing methods, achieving an average classification accuracy of 76.9% for within-subject and 68.8% for cross-session tasks across five publicly available datasets. In within-subject scenarios, it surpasses competing methods by 2.6% to 14.6% (p < 0.001), while in cross-session scenarios, it outperforms others by 2.4% to 13.8% (p < 0.05). Visualized filters confirm the alignment with motor cortex neural mechanisms. Conclusion: CTSSP provides an effective solution to extract dynamic temporal-spectral-spatial features, improving classification accuracy in MI-BCI tasks. Significance: The CTSSP algorithm offers a robust, interpretable method for MI-BCI systems, enhancing performance in noisy, non-stationary environments and advancing neuro-rehabilitation applications.

20 January 2026 in Journal of Neural Engineering
Verified
Journal Article
Closed access
LPLincong Pan
Lincong Pan
KWKun Wang
Kun Wang
WYWeibo Yi
Weibo Yi
YZYang Zhang
Yang Zhang
MXMinpeng Xu
Minpeng Xu
DMDong Ming
Dong Ming
Objective. Motor imagery brain–computer interfaces hold significant promise for neurorehabilitation, yet their performance is often compromised by electroencephalography (EEG) non-stationarity, low signal-to-noise ratios, and severe cross-session variability. Current decoding methods typically suffer from fragmented optimization, treating temporal, spectral, and spatial features in isolation. Approach. We propose common temporal–spectral-spatial patterns (CTSSP), a unified framework that jointly optimizes filters across all three domains. The algorithm integrates: (1) multi-scale temporal segmentation to capture dynamic neural evolution, (2) channel-adaptive finite impulse response filters to enhance task-relevant rhythms, and (3) low-rank regularization to improve generalization. Main results. Evaluated across five public datasets, CTSSP achieves state-of-the-art performance. It yielded mean accuracies of 76.9% (within-subject), 68.8% (cross-session), and 69.8% (cross-subject). In within-subject and cross-session scenarios, CTSSP significantly outperformed competing baselines by margins of 2.6%–14.6% ( p < 0.001) and 2.3%–13.8% ( p < 0.05), respectively. In cross-subject tasks, it achieved the highest average accuracy, proving competitive against deep learning models. Neurophysiological visualization confirms that the learned filters align closely with motor cortex activation mechanisms. Significance. CTSSP effectively overcomes the limitations of decoupled feature extraction by extracting robust, interpretable, and coupled temporal–spectral-spatial patterns. It offers a powerful, data-efficient solution for decoding MI EEG in noisy, non-stationary environments. The code is available at https://github.com/PLC-TJU/CTSSP .

23 August 2024
Verified
Conference Paper
Open access
LPLincong Pan
Lincong Pan
KWKun Wang
Kun Wang
XSXinwei Sun
Xinwei Sun
MXMinpeng Xu
Minpeng Xu
DMDong Ming
Dong Ming
Motor imagery (MI) is a mental process without actual movement, which can be recognized by electroencephalogram (EEG). Effective pre-processing and feature extraction methods are essential for improving the performance of MI-EEG classification. This paper proposes a novel Riemannian geometry-based spatial filtering (RSF) method, which projects the MI-EEG signals onto a lower-dimensional subspace that maximizes the Riemannian distance between the covariance matrices of different classes. The paper evaluates the performance of RSF on a public dataset (Pan2023) and applies it to several common classification algorithms, such as common spatial pattern (CSP), filter bank CSP (FBCSP), tangent space linear discriminant analysis (TS-LDA), ShallowNet, DeepNet, EEGNet and Graph-CSPNet. The results show that RSF can significantly improve the classification accuracy of these algorithms, and reduce the subsequent computation time cost, especially for those neural network algorithms with complex structures. These results indicate that RSF is an efficient spatial filtering method for MI-EEG classification, and provides a new perspective and method for EEG-based applications.