(Press-News.org)
Advancing high-speed steady-state visually evoked potential (SSVEP)-based brain–computer interface (BCI) systems require effective electroencephalogram (EEG) decoding through deep learning. SSVEP-BCIs enable direct communication between the brain and external devices (e.g., spellers, prosthetics) by detecting EEG signals triggered by visual stimuli flickering at specific frequencies. They are prized for their high information transfer rate (ITR)—a key measure of BCI speed—and minimal training needs. “However, two critical barriers have hindered their performance: Data sparsity: Collecting large EEG datasets is time-consuming, costly, and constrained by subject availability, leading to overfitting in deep learning models. Signal complexity: Traditional models (e.g., CNNs) struggle to capture the temporal, spatial, and frequency features of dynamic EEG signals, limiting decoding efficiency—especially with short training data.” explained study author Jin Yue.
To overcome these hurdles, we developed two complementary technologies: 1. Background EEG Mixing (BGMix): Neural-Inspired Data Augmentation Unlike generic data mixing techniques (e.g., Mixup), BGMix leverages the neurophysiology of EEG signals: SSVEP signals consist of stable task-related components (linked to visual stimuli) and variable background noise (unrelated to the task). BGMix generates new training samples by swapping background noise between EEG trials of different classes—preserving critical task features while introducing natural variability. This approach ensures augmented samples retain the statistical distribution of real EEG data, avoiding the "unnatural" signals produced by conventional methods. 2. Augment EEG Transformer (AETF): Multidimensional Feature Capture The team paired BGMix with AETF, a Transformer-based model designed specifically for EEG decoding. AETF integrates three key modules to extract comprehensive features: A fully connected layer for spatial filtering (capturing signal differences across electrodes). A convolutional layer for frequency filtering (targeting SSVEP’s stimulus-linked frequency patterns). A 2-layer Transformer encoder for temporal feature extraction (preserving long-range timing relationships critical for fast BCI responses). Unlike CNNs, which lose temporal information via pooling, AETF’s attention mechanism focuses on critical time points—making it ideal for short training data.
The study’s innovations address a critical gap in BCI usability: high-speed performance with limited data. For users (e.g., individuals with motor disabilities), this means faster, more reliable communication and device control. “To further enhance real-world applicability, our team explored model compression via knowledge distillation: a simplified AETF variant (AETF_1layer) retained 95–98% of the original model’s performance while reducing computational cost—critical for portable BCI devices.” emphasized the authors. Looking ahead, we plan to integrate multimodal data (e.g., eye tracking) and transfer learning to expand AETF’s utility across more BCI tasks and subject groups.
Authors of the paper include Jin Yue, Xiaolin Xiao, Kun Wang, Weibo Yi, Tzyy-Ping Jung, Minpeng Xu, and Dong Ming.
This work was supported by the the STI 2030-Major Projects 2022ZD0210200 and the National Natural Science Foundation of China (nos. 82330064, 62106170, and 62006014).
The paper, “Augmenting Electroencephalogram Transformer for Steady-State Visually Evoked Potential-Based Brain–Computer Interfaces” was published in the journal Cyborg and Bionic Systems on Oct 07, 2025, at DOI: 10.34133/cbsystems.0379.
END
The competition is fierce. Olympic gymnast Simone Biles is pushing herself to perform stronger and more consistently. Then, her coach calls her “fat.” It’s meant as a motivation, but this time, it has the opposite effect.
While some coaches might believe such comments could spur an athlete to train harder, new research involving the Binghamton University School of Management found a better way to boost an athlete’s mental toughness: transformational leadership techniques focusing on self-determination. These techniques have ...
"Recent advancements in neuromorphic computing have enabled significant progress in dynamic motion recognition, yet distinguishing high-speed and low-speed movements remains computationally challenging due to the limited dynamic range of conventional CMOS technology. Our work introduces a SnS₂-based in-sensor reservoir computing device that leverages tunable multi-timescale optoelectronic dynamics to address this bottleneck," explained study author Linfeng Sun, a professor at Beijing Institute of Technology. ...
T cells play central roles in the adaptive immune response against cancer. Their functional inactivation is a primary driver of tumor progression, making the reactivation of T cell function a main goal in immunotherapy. The review details how T cells specifically recognize and eliminate malignant cells by engaging tumor antigen peptides presented by MHC molecules.
“T cells can specifically recognize tumor antigen epitopes presented by MHC molecules to clear malignant tumor cells. The targets that can be recognized by T cells in tumors mainly include two types”, the authors point out. Highly immunogenic, tumor-specific "neoantigens" derived from somatic ...
Rising seas are irreversible on human time scales and among the most severe consequences of climate change. Emissions released in the coming decades will determine how much coastlines are reshaped for centuries to come. New research shows that near-term mitigation could spare future generations around 0.6 meters of sea-level rise that would be caused by emissions between 2020 and 2090 following current policies, making today’s decisions critical not only for limiting warming but also for coastal ...
Findings from Mass General Brigham investigators highlight the intricate interplay of diverse physiological processes as the brain shifts from wakefulness to sleep
A new study by investigators from Mass General Brigham used next-generation imaging technology to discover that when the brain is falling asleep, it shows a coordinated shift in activity. They found that during NREM (non-rapid eye movement) sleep, parts of the brain that handle movement and sensory input stay active and keep using energy, while ...
In the battle against mosquito-borne diseases that kill hundreds of thousands of people each year, scientists turned to an unlikely ally: a fungus that smells like flowers.
Taking advantage of the mosquito’s natural attraction to flowers, an international team of researchers engineered a new strain of Metarhizium fungus that imitates a flower’s sweet scent and lures the bloodsucking bugs to their deaths.
Inspired by certain fungi that the team found emit a sweet-smelling chemical called longifolene to draw ...
Researchers have achieved a breakthrough in solar physics by providing the first direct evidence of small-scale torsional Alfvén waves in the Sun's corona – elusive magnetic waves that scientists have been searching for since the 1940s.
The discovery, published today in Nature Astronomy, was made using unprecedented observations from the world's most powerful solar telescope, the U.S. National Science Foundation (NSF) Daniel K. Inouye Solar Telescope in Hawaii.
The findings could finally explain one of the Sun's greatest mysteries ...
Winter vomiting disease is caused by the Norovirus, which is most virulent during the colder half of the year. The infection clears up after a couple of days, but the protection it provides is short-lived, meaning that the same person can fall repeatedly sick in a short space of time. But some people cannot succumb to the virus, thanks to a particular gene variant.
“We wanted to trace the historical spread of the gene variant,” says Hugo Zeberg, senior lecturer in genetics at the Department of Physiology and Pharmacology, Karolinska Institutet, and researcher at the Max Planck Institute for Evolutionary Anthropology in Leipzig.
Defective gene protects against ...
New research finds that common nervous system components are present in the gastrointestinal tract where they influence tumour growth, revealing untapped opportunities for cancer treatment.
GLOBAL: Australian researchers have identified two nervous system components that drive tumour growth in gastrointestinal cancers, creating promising new avenues for treatment with existing approved therapies.
Our gut contains its very own nervous system and is commonly regarded as the second brain. Key players of this system are neuropeptides, the signalling ...
More expensive steak, cheaper tomatoes, but the same total cost for the average basket of groceries at the supermarket. A comprehensive study, led by researchers from Chalmers University of Technology in Sweden has analysed the potential effects of a food tax shift – where VAT is removed from healthy foods and levies are introduced on foods that have a negative impact on the climate. The study shows that a shift in taxes could have both environmental and human health benefits, and means that 700 fewer people in Sweden would die prematurely each year.
Today, diet in many high-income countries is a leading risk factor for certain diseases and premature ...
