Brain-computer interfaces: the innovative key to unlocking neurological conditions

The paper addresses the growing global burden of neurological diseases, including stroke, Parkinson’s disease, and psychiatric disorders, which profoundly impact mortality, disability, quality of life, and socio-economic costs. As populations age, incidence is projected to rise. Traditional treatments often fail to adequately address complex neurological dysfunctions, highlighting an urgent need for innovative strategies. Brain-Computer Interface (BCI) technology enables bidirectional communication between brain activity and external devices, translating neural signals into actionable commands or delivering feedback to the brain. BCIs offer diagnostic advantages through noninvasive monitoring and deep insights into brain function; therapeutic potential by restoring lost functions via prosthetics or computer control; and rehabilitation through activation of specific neural circuits to promote neuroplasticity. Integrating artificial intelligence and machine learning improves precision, adaptability, and personalization by learning individual brain patterns. BCIs are inherently interdisciplinary, involving neurosurgery (for electrode implantation), biomedical engineering, neuroscience, computer science, AI, and materials science to ensure safety and stability. This review synthesizes the developmental history of BCI, applications across neurological conditions, and future directions, emphasizing clinical translation and neurosurgery’s pivotal role.

This review article compiles and synthesizes recent advances in BCI technology across multiple neurological domains. It outlines the evolution of BCI in three phases—Academic Exploration (1924–1970), Scientific Validation (1970–2000), and Experimental Application (2001–present)—highlighting key milestones (e.g., Hans Berger’s discovery of brain waves, Vidal’s BCI concept, early human brain chips, brain-to-brain interfaces, bidirectional closed-loop BCIs). It classifies BCIs by invasiveness (noninvasive: EEG, fMRI, fNIRS, MEG; semi-invasive: ECoG; invasive: SEEG) and by directionality (unidirectional vs. bidirectional). The review details the working principles (signal acquisition, preprocessing, feature extraction, classification with SVM/LDA/ANN/DL, device control and feedback), and surveys common paradigms (motor imagery, SSVEP, AEP, P300). It provides neuroscience fundamentals pertinent to BCI (CNS/PNS structures, neuronal signaling, synaptic transmission). The article then examines clinical and research applications: motor disturbances (PD, stroke, SCI, locked-in syndrome, epilepsy), disorders of consciousness, cognitive/mental disorders (AD, depression), and sensory disorders (auditory, vision, speech). It references randomized controlled trials and technological advances, with supplementary tables for movement disorders and cognitive/mental disorders, and figures illustrating classifications and applications. Emerging themes include AI integration, closed-loop systems, bidirectional interfaces, and ethical and safety considerations.

This is a comprehensive descriptive review synthesizing contemporary BCI research and clinical studies across multiple neurological conditions. The authors provide an updated overview of BCI classifications, operational principles, paradigms, and neuroscience fundamentals, followed by critical appraisals of applications in movement, consciousness, cognitive/mental, and sensory disorders. Evidence cited includes randomized controlled trials, cohort studies, case studies, and technological demonstrations. Figures visualize development timelines and system workflows; supplementary tables summarize RCTs in movement disorders and cognitive/mental disorders. No systematic search strategy, inclusion/exclusion criteria, or meta-analytic methods are reported; the review focuses on narrative synthesis, technological trends, and prospective directions.

Highlights and representative findings across domains:

• Parkinson’s disease (PD) and DBS: RCTs show DBS improves motor symptoms and increases 'on' time versus best medical therapy; in early PD, DBS plus medication improved quality of life (PDQ-39 changed by +7.8 points vs −0.2 control), reduced levodopa equivalent dose by 39% (vs +21% control), and enhanced daily living activities; STN-DBS at 80 Hz improved assembly task performance; closed-loop DBS leveraging subthalamic beta oscillations shows promise, with parameter selection and motor state affecting adaptive DBS efficacy.
• Stroke: In an RCT of 28 subacute patients, BCI-augmented motor imagery yielded superior functional recovery compared with conventional motor imagery; EEG revealed enhanced ipsilateral alpha/beta desynchronization. A hybrid EEG/EOG BCI enabled integrated wheelchair and robotic arm control; in testing (22 subjects), five completed a complex beverage task, demonstrating feasibility of multi-device control. BCIs leveraging theta/alpha neurofeedback improved memory encoding and cognitive function.
• Spinal cord injury (SCI): Multimodal BCIs enhanced lower limb rhythmic movement and alleviated pain; BCI-assisted motor imagery improved upper limb function, especially early post-injury; implanted microelectrode arrays enabled precise bionic limb control, improving autonomy.
• Locked-in syndrome: Eye-tracking integrated with HMM/DNN improved character recognition and input speed; fMRI-guided ECoG placement increased effectiveness; standardized methodologies for implantable communication BCIs were proposed.
• Epilepsy: Closed-loop stimulation (e.g., RNS) associated with improved seizure control via long-term network modulation; a wireless neural prosthesis (ECoGIW-16E) supported up to 6 months of ECoG recording and direct cortical stimulation in primates; memristor-based analog neural signal analysis achieved 93.46% accuracy with ~400× power efficiency over leading CMOS.
• Disorders of consciousness (DoC): Behavioral diagnostics risk up to 43% misdiagnosis of VS; fMRI and EEG-based BCIs detect residual volitional activity and covert consciousness; ERPs and automated classification differentiate VS from MCS; P300 spellers enabled limited communication in MCS.
• Alzheimer’s disease (AD): EEG biomarkers (e.g., increased theta power) predicted cognitive decline and supported early detection; neurofeedback stabilized or improved memory/attention; conditioning-based BCI linked yes/no thoughts to emotional stimuli for basic communication; exploratory photon emission monitoring from hippocampus proposed for minimally invasive diagnostics.
• Depressive disorder: Subcallosal cingulate DBS was safe and feasible; sham-controlled trial showed no significant difference in response rates over 6 months; preclinical NAc-DBS reduced depressive-like behaviors via GABA modulation; open-label HB-DBS showed ~49% symptom reduction at 1 month with sustained improvements; EEG-based BCIs (ResNets) classified and scored depression emphasizing beta-band features; BCI-enabled psychoneurotherapy reduced high-beta activity and depressive symptoms.
• Sensory disorders: Auditory BCI achieved up to 95% detection accuracy using auditory stream segregation; P300 auditory BCIs supported communication in ALS; frequency-specific beep paradigms outperformed musical notes; hybrid ASSR+P300 improved accuracy and stability. Vision: high-frequency SSVEP with computer vision enabled robotic arm control for object manipulation. Speech: chronic ECoG-based RNN synthesis produced intelligible speech in ALS; high-performance speech neuroprostheses decoded commands stably over 3 months without recalibration; long-term implanted BCIs enabled reliable device control.
• Anesthesia/intraoperative awareness: MNS-based BCI improved classification of motor imagery during general anesthesia; under propofol, offline classifiers detected movement attempts with accuracies up to 85% and 83%, including cross-state detection from awake-trained models.
• Emerging trends: Integration of AI/ML for adaptive decoding; development of bidirectional closed-loop systems; emphasis on personalization, remote monitoring, and standardized protocols.

The synthesis demonstrates BCIs’ broad potential to address diagnostic gaps, restore functions, and enhance rehabilitation across diverse neurological conditions. Objective neuroimaging and EEG-based BCIs improve detection of covert consciousness and early cognitive decline, complementing traditional assessments. In movement disorders and SCI, motor imagery BCIs with real-time feedback facilitate neuroplasticity, improving motor recovery and autonomy; bidirectional interfaces add sensory feedback to refine control. In epilepsy, closed-loop neuromodulation and energy-efficient neural analysis advance seizure management. For AD and depression, BCIs enable early diagnostics, neurofeedback-based symptom modulation, and targeted neuromodulation, while speech BCIs restore communication for severe motor and speech impairment. AI and deep learning improve decoding from complex signals, personalize interfaces, and maintain stability across states. Neurosurgery underpins clinical translation through precise implantation and integration of semi- and fully invasive systems. The findings underscore ethical imperatives—data privacy, identity, autonomy—and the need for standardized protocols and long-term efficacy data to bridge laboratory success and routine clinical practice.

BCI technology offers transformative opportunities in diagnosing, treating, and rehabilitating neurological disorders, yet widespread clinical deployment remains constrained by technical, ethical, and translational challenges. Priorities include enhancing biocompatibility and long-term stability of implants, miniaturization and portability, robust real-time decoding algorithms, intuitive user interfaces, and rigorous validation of clinical efficacy. Future work should emphasize bidirectional and high-performance BCIs, long-term stable implantable electrodes, closed-loop stimulation systems, and large-scale clinical trials, alongside robust ethical frameworks and international standardization to promote interoperability and data sharing. Interdisciplinary collaboration—especially neurosurgery—will be crucial in translating BCI advances into practice. With responsible development and continued innovation, BCIs are poised to substantially improve quality of life and redefine neurotechnology’s role in healthcare.

The review is narrative and does not report a systematic search strategy, inclusion criteria, or quantitative synthesis. Many applications remain experimental with limited long-term clinical data. Key constraints include implant biocompatibility and stability, decoding accuracy and generalizability across conditions and states, system miniaturization and portability, user interface usability, and safety. There is a lack of standardized protocols, scarce large-scale randomized clinical trials, and challenges in bridging laboratory findings to routine clinical practice. Ethical and privacy issues surrounding neural data and human–machine integration require robust safeguards.