Applications of functional ultrasound brain-computer interfaces in neurological diseases: research progress based on...

Brain-computer interfaces (BCIs) bypass peripheral nerves and muscles to establish direct information pathways between the brain and external devices, providing novel approaches for motor function reconstruction, communication assistance, and neurorehabilitation in patients with neurological disorders. However, conventional non-invasive BCIs relying on modalities such as electroencephalography (EEG), magnetoencephalography (MEG), or functional magnetic resonance imaging (fMRI) exhibit limitations in spatiotemporal resolution, portability, or real-time capability. Functional ultrasound (fUS), an emerging neuroimaging technology based on neurovascular coupling (NVC), detects cerebral blood volume changes at a spatial resolution of approximately 100 μm; its imaging frame interval is 200-500 ms, whereas its effective temporal resolution, limited by the hemodynamic response, is 1-2 s, offering a potential solution to overcome these constraints. This review comprehensively analyzed NVC-based fUS brain-computer interfaces from the dimensions of technical principles, clinical applications, future directions, and challenges. fUS has demonstrated significant clinical value in stroke assessment, epilepsy monitoring, and early identification of neurodegenerative diseases. Emerging technologies such as transcranial focused ultrasound neuromodulation, sonogenetics, and intravascular ultrasound are driving the evolution of fUS from single-signal "acquisition" toward bidirectional "acquisition-modulation" closed-loop systems, which remain at the proof-of-concept stage. However, the hemodynamic response latency relative to neural electrical activity (1-2 s), the yet-unclear biosafety of chronic ultrasound exposure, and insufficient decoding reliability under pathological conditions remain critical bottlenecks requiring urgent resolution. In conclusion, NVC-based fUS brain-computer interfaces provide a novel tool for precision diagnosis and treatment of neurological diseases. Future efforts should focus on conducting multicenter clinical trials and enhancing the robustness of decoding algorithms.