A groundbreaking study led by scientists at Tianjin University has unveiled a revolutionary noninvasive technique to locate and decode intracranial endogenous signals with remarkable spatiotemporal precision. This advancement is expected to propel the field of brain–computer interfaces forward by overcoming the long-standing limitations inherent in conventional electroencephalography (EEG) technologies. Published in the journal Cyborg and Bionic Systems on April 9, 2025, the research details an innovative fusion of transcranial focused ultrasound and EEG signal modulation, promising new horizons for high-resolution neuroimaging and brain signal interpretation.
Traditional scalp EEG, while widely used for monitoring brain activity, has been fundamentally constrained by its low spatial resolution. This limitation is rooted in the volume conduction effect, where electrical signals diffuse through intervening tissues such as the scalp, skull, and cerebrospinal fluid, resulting in blurred and spatially imprecise measurements. These constraints have posed considerable challenges for applications requiring fine-grained neural decoding, such as advanced brain–computer interface (BCI) systems that demand pinpoint accuracy in identifying neuronal intent.
To combat these challenges, the team from Tianjin University developed a sophisticated transcranial focused ultrasound (tFUS) phased-array system designed to modulate EEG signals and thereby enhance their spatial and temporal resolution. Employing a 128-element phased array transducer, the researchers established a numerical simulation framework and an experimental platform based on real anatomical brain models. This approach aided in the accurate modeling of ultrasonic wave propagation and its interaction with complex skull structures, which were previously oversimplified or neglected in existing studies.
A core innovation of this study involves the establishment of a three-dimensional transcranial multisource dipole localization and decoding system. Unlike traditional two-dimensional models, this approach accounts for the intricate anatomical heterogeneities of the skull and brain tissue. It integrates clues from acoustoelectric coupling phenomena, where the mechanical energy of ultrasound waves modulates electrophysiological signals. This hybrid modality thus bridges the gap between ultrastructural acoustic focusing and neuronal electrical activity, enabling superior signal extraction fidelity.
Prior research into ultrasound-modulated EEG (USMEEG) often sidelined the influence of realistic skull morphology and frequently relied on envelope-based signal processing algorithms with unstable performance. Addressing these limitations, the Tianjin University team introduced a pulse repetition frequency (PRF) sideband algorithm, substantially boosting the robustness and resolution of intracranial signal localization and decoding. This innovation leverages the unique temporal frequency characteristics imparted by ultrasound modulation, allowing enhanced discrimination of neural signal sources amidst noise.
Experimentally, the researchers validated their phased-array focused ultrasound platform both in pure water environments and transcranial scenarios. The system demonstrated precise focusing capabilities with a modulation range confined to approximately 10 millimeters, well within safe acoustic exposure thresholds. Focal acoustic pressures achieved were more than 200% higher than those attainable via conventional self-focusing transducers, showcasing the substantial amplification benefits afforded by phased-array steering combined with ultrasonic modulation.
Quantitative analysis of dipole localization revealed a dramatic improvement in signal-to-noise ratio (SNR). The newly developed algorithm attained an SNR of 24.18 decibels, marking a 50.59% increase over traditional envelope detection methods. Moreover, source signal decoding accuracy consistently exceeded 85%, underscoring the method’s potential for reliable extraction of brain signals critical for neuroprosthetic control, neurological diagnostics, and fundamental neuroscience research.
This research presents an experimental and computational foundation that paves the way for noninvasive high-spatiotemporal-resolution EEG acquisition technologies. Such platforms could redefine brain–computer interfaces by providing unprecedented insight into neural dynamics without the risks and complexities associated with invasive implants. According to Hao Zhang, the lead author and researcher at Tianjin University, this breakthrough offers much-needed technical support for precise brain–computer manipulation applications, ranging from neurorehabilitation to cognitive enhancement.
Beyond the direct technological contributions, this study’s comprehensive modeling approach represents a paradigm shift in how transcranial ultrasound and electrophysiological signals are integrated. By incorporating realistic skull and brain models and extending simulations and experiments into three-dimensional spaces, the team’s work transcends idealized laboratory settings, approaching conditions more representative of in vivo human applications. This aspect dramatically enhances the translational impact of the findings.
Future implications of this work are vast. Improved intracranial source localization through USMEEG technology could accelerate the decoding of complex brain states, enabling seamless interaction between humans and machines. Potential areas of expansion include personalized neuromodulation therapies, early detection of neurological disorders, and sophisticated closed-loop BCI systems capable of real-time adaptive modulation based on precise brain signal feedback.
The team behind these innovations comprises Hao Zhang, Xue Wang, Guowei Chen, Yanqiu Zhang, Xiqi Jian, Feng He, Minpeng Xu, and Dong Ming. Their multidisciplinary expertise spans ultrasonic engineering, computational neuroscience, signal processing, and biomedical instrumentation, which collectively contributed to the success of this multifaceted project.
Funding and support were provided by prominent Chinese science foundations and innovation programs, including the National Key Research and Development Program of China, the National Natural Science Foundation, various postdoctoral grants, and specialized brain–computer interaction laboratories. This extensive backing affirms the strategic importance of research efforts targeting noninvasive neurotechnologies with clinical and technological significance.
With its publication in Cyborg and Bionic Systems, this cutting-edge research enters the global neuroscience and neuroengineering communities’ spotlight. The study’s methods, results, and open challenges provide a rich template for future investigations aiming to harness acoustoelectric coupling and ultrasonics to unlock the brain’s complexities. Integrated multidisciplinary collaborations and continued technological refinements will likely energize this promising domain.
As brain–computer interfaces continue to evolve, Tianjin University’s contributions set a new benchmark for what is achievable in neural decoding without compromising safety or requiring invasive procedures. The demonstrated improvements in spatial resolution, decoding accuracy, and focal control herald a future where seamless mind-machine integration could become an everyday reality, reshaping fields from medicine to entertainment and beyond.
Subject of Research: Noninvasive intracranial source signal localization and decoding using transcranial focused ultrasound modulated EEG with high spatiotemporal resolution.
Article Title: Noninvasive Intracranial Source Signal Localization and Decoding with High Spatiotemporal Resolution
News Publication Date: April 9, 2025
Web References: DOI: 10.34133/cbsystems.0206
Image Credits: Hao Zhang, Tianjin University
Keywords: Applied sciences and engineering; Health and medicine; Life sciences
