Abstract Brain – computer interfaces (BCIs) require rapid and accurate decoding of neural activity, yet conventional computing architectures face growing latency as neural recording scales. We demonstrate a quantum computing – enabled neural decoding approach using a physical 1000-qubit coherent photonic Ising machine, in which inference is performed through hardware energy relaxation rather than numerical computation. By mapping sparse neural spike patterns onto Ising Hamiltonians, our hardware-native Quantum Semi-Restricted Boltzmann Machine achieves up to 96.2% accuracy across public in vivo datasets spanning multiple species and modalities. We report hardware-verified median latencies of 0.075 ms— a tenfold speedup over GPUs—with complexity-invariant scaling. These results establish quantum computing as a viable pathway toward ultra-low-latency neural decoding for future BCI systems.

Neural electrodes face persistent challenges, including inflammation, biofouling, and impedance, which compromise long-term recording and stimulation quality. Here, we introduced a multifunctional polyamino acid interface that enhances neural interfacing by combining biocompatibility, antimicrobial activity, and antifouling properties. Applied to flexible electrodes, this coating reduces foreign body reaction and preserves neuronal proximity, ensuring stable integration with brain tissue. In chronic rodent models, functionalized electrodes achieve high-fidelity single-unit recordings for over 300 days with significantly higher spike amplitudes and yield than bare controls. Notably, the interface enables an order-of-magnitude improvement in neuromodulation efficiency, evoking robust motor responses at only 2 µA compared to 100 µA for uncoated probes. Multi-omics analysis reveals a molecularly profound alteration in the host tissue response, transitioning from a pro-inflammatory injury signature to an attenuated inflammatory state with improved tissue homeostasis. Furthermore, the interface's resistance to biological "cementing" facilitates damage-free electrode removal and recovery, preserving the neural architecture and enabling the possibility of chronic replacement. This universal platform offers a promising and practical strategy for the next generation of stable, clinically viable neural-computer interfaces.

High-density, long-term stable decoding of whole-brain function is crucial for advancing basic neuroscience research and developing neural disorder therapies. However, two major challenges remain: the lack of scalable interfaces capable of long-term, multi-regional recordings and the limited generalizability of existing decoding algorithms across days and individuals. Here, we developed an integrated platform that achieves accurate, stable, and generalizable decoding of behavioral states (resting, roaming, feeding, and flash) with up to 89% accuracy. This platform combines multi-region flexible probes (MRFPs), enabling distributed recordings from 128 sites across eight brain regions over months, with a Conformer-based deep learning framework optimized for brain-wide neural dynamics. Comparative analyses demonstrate that distributed sampling, particularly from five or more regions, markedly enhances decoding performance over concentrated electrode configurations. Furthermore, the platform supports robust generalization across days and individuals without retraining, providing a practical solution for longitudinal and large-scale behavioral neuroscience studies. These results establish a foundation for stable, high-fidelity multi-region electrophysiology and offer a generalizable approach for decoding internal states from complex neural dynamics.

Near-infrared (NIR) light offers opportunities for minimally invasive deep-tissue neuromodulation; however, its photon energy is insufficient to directly activate light-sensitive agents. Upconversion nanoparticles (UCNPs) address this limitation by converting NIR to visible light, enabling precise neuromodulation. Beginning with the fundamental properties of UCNPs, including upconversion mechanisms, brightness enhancement strategies, photothermal behaviors, and biocompatibility, this review then highlights UCNP-enabled neuromodulation strategies: optogenetic control, photochemical modulation, optoelectronic neural stimulation and visual activation. By bridging materials innovation and neuroengineering applications, this review positions UCNPs as transformative tools for dissecting neural circuits, treating neurological disorders, and advancing minimally invasive brain-machine interfaces.

This paper reports a Hybrid Surface-Deep Flexible Array (HSDFA) enabling simultaneous cross-scale neural recordings in freely moving mice. This design addresses the limitations of existing separate setups, which suffer from placement mismatch and relative drift. The device integrates 60 conformal surface electrodes and 68 intracortical electrodes on a single polyimide substrate, ensuring intrinsic spatial alignment between surface and deep recording sites. The average impedance of PEDOT:PSS modified electrodes is reduced to 82 ± 19 kΩ (surface) and 378 ± 52 kΩ (depth) at 1 kHz, ensuring high-fidelity signal acquisition. We applied the HSDFA to the simultaneous recording of low-frequency waves on the external skull surface and intracortical Local Field Potentials (LFPs), revealing the high-spatiotemporal propagation across the surface and deep brain. By analyzing the cross-scale coupling across different frequencies, we found that the correlation between deep and surface signals is depth- and frequency-dependent, with the coherence increasing at shallower depths and lower frequency bands. The HSDFA provides a powerful tool for investigating cross-scale neural dynamics.

In this article, we present a silk-based bidirectional and flexible extravascular bio-interface designed for the monitoring and regulation of vascular electrical activity, which is closely associated with sympathetic nervous system. By incorporating bioadhesive and biocompatible silk fibroin materials, the biointerface exhibits waterproof adhesive properties and conforms seamlessly to the curved, three-dimensional surface of blood vessels without the need for additional fixation tools. The supreme electrical properties of the extravascular biointerface facilitate reliable, high-resolution electrophysiological signal acquisition and effective neural stimulation, which is validated via in vivo experiments. This novel bidirectional, flexible extravascular biointerface holds potential for advancing clinical electronic medicine for cardiovascular disease management.

Neuromodulation is crucial for advancing neuroscience and treating neurological disorders. However, traditional methods using rigid electrodes have been limited by large stimulating currents, low precision, and the risk of tissue damage. In this work, we developed a biocompatible ultraflexible electrode array that allows for both neural recording of spike firings and low-threshold, high-precision stimulation for neuromodulation. Specifically, mouse turning behavior can be effectively induced with approximately five microamperes of stimulating current, which is significantly lower than that required by conventional rigid electrodes. The array's densely packed microelectrodes enable highly selective stimulation, allowing precise targeting of specific brain areas critical for turning behavior. This low-current, targeted stimulation approach helps maintain the health of both neurons and electrodes, as evidenced by stable neural recordings after extended stimulations. Systematic validations have confirmed the durability and biocompatibility of the electrodes. Moreover, we extended the flexible electrode array to a brain-to-brain interface system that allows human brain signals to directly control mouse behavior. Using advanced decoding methods, a single individual can issue eight commands to simultaneously control the behaviors of two mice. This study underscores the effectiveness of the flexible electrode array in neuromodulation, opening new avenues for interspecies communication and potential neuromodulation applications.

In this study, we present a high-density, high-channel-count micro-electrocorticographic (mu ECoG) electrode array for real-time motor decoding. The 256-channel mu ECoG electrode arrays, based on MEMS processes, possess excellent flexibility and mechanical robustness, allowing them to conform to the cortical surface and enabling the acquisition of high-quality ECoG signals. The advanced brain-computer interface (BCI) system was applied to a Labrador dog and high accurate real-time motor decoding was achieved, showing the advantages of high-resolution ECoG sampling. Our method demonstrates the potential for controlling a cursor with the ECoG signals, offering the possibility of reconstructing motor functions and synthesizing avatars.

We report a flexible silk-based stent-electrode system designed for minimally invasive vascular implantation with controllable drug release, featuring vasospasm monitoring and real-time drug delivery capabilities. Comprehensive characterization of its mechanical properties, electrical performance, and drug release functionality has been conducted, with the feasibility of the system successfully demonstrated in vitro. The silk-based stent exhibits outstanding mechanical strength and biocompatibility, enabling precise control over drug release. The electrode array, constructed on a flexible PI substrate, is produced using a double-sided MEMS process, ensuring conformal implantation and optimal integration within vascular environments. Therefore, our vascular stent-electrode system demonstrates great potential for monitoring vascular electrophysiological activity, detecting vasospasm, and providing real-time therapeutic drug delivery for treating vascular-related diseases.

Silk fibroin, recognized for its biocompatibility and modifiable properties, has significant potential in bioelectronics. Traditional silk bioelectronic devices, however, face rapid functional losses in aqueous or in vivo environments due to high water absorption of silk fibroin, which leads to expansion, structural damage, and conductive failure. In this study, we developed a novel approach by creating oriented crystallization (OC) silk fibroin through physical modification of the silk protein. This advancement enabled the fabrication of electronic interfaces for chronic biopotential recording. A pre-stretching treatment of the silk membrane allowed for tunable molecular orientation and crystallization, markedly enhancing its aqueous stability, biocompatibility, and electronic shielding capabilities. The OC devices demonstrated robust performance in sensitive detection and motion tracking of cutaneous electrical signals, long-term (over seven days) electromyographic signal acquisition in live mice with high signal-to-noise ratio (SNR >20), and accurate detection of high-frequency oscillations (HFO) in epileptic models (200-500 Hz). This work not only improves the structural and functional integrity of silk fibroin but also extends its application in durable bioelectronics and interfaces suited for long-term physiological environments.

Anti-seizure medications and deep brain stimulation are widely used therapies to treat seizures; however, both face limitations such as resistance and the unpredictable nature of seizures. Recent advancements, including responsive neural stimulation and on-demand drug release, have been developed to address these challenges. However, a gap remains, as electrical stimulation provides only transient effects while medication has a delayed onset. To bridge this gap, we developed a Bimodal Closed-loop Neurostimulation Implant System that integrates real-time neural recording, immediate electrical stimulation, and on-demand drug release to achieve more effective seizure suppression. This dual-modality system combines rapid electrical intervention with sustained pharmacological treatment to provide comprehensive seizure control. An embedded platform powered by a Long Short-Term Memory network detects seizures and autonomously triggers these interventions. In vivo studies in an epileptic mouse model revealed that electrical stimulation achieved rapid seizure suppression, terminating 75.16% of seizures, with 90% of episodes suppressed within 10 s. The subcutaneous drug capsule provided additional control, with an onset of action approximately 15 min after release. The dual-modality approach bridged the gap between immediate and delayed intervention, stabilizing neural activity and reducing seizure recurrence. Furthermore, we confirmed the long-term viability of neurons, observing no significant changes in morphology or signal quality following stimulation and drug release. These results suggest that the system offers rapid, stable, and minimally invasive seizure control, making it a promising therapeutic tool for epilepsy. By bridging the gap between electrical effects and delayed pharmacological action, the system presents a novel approach to epilepsy management.

Optoelectronic neuromodulation has transformed neuroscience research and holds great promise for treating neurological disorders. However, conventional optoelectronic methods rely on ultraviolet/visible light, which poorly penetrates tissue and typically necessitates surgically implanted optical fibers for deep‐brain stimulation. Here, a heterostructure is presented that integrates near‐infrared (NIR)‐excitable upconversion nanoparticles (UCNPs) and broadband‐absorbing CsPbBr 3 perovskite quantum dots (QDs). This nanostructure converts deeply penetrating 980 nm NIR light into localized electrical stimuli, enabling immediate and precise modulation of neuronal activity without implants. In vitro, NIR illumination of this heterostructure reliably increases the firing rate of wild‐type dopaminergic (DA) neurons in acute brain slices. Importantly, in vivo, transcranial NIR stimulation of the heterostructure in the secondary motor cortex (M2) and ventral tegmental area (VTA) modulates neuronal activity, triggers turning behavior, and promotes dopamine release. Moreover, it exhibits negligible neuroinflammation and structural stability in brain tissue over at least four weeks. By integrating a stable heterostructure for efficient NIR‐driven photocurrent generation, the method offers a non‐genetic, minimally invasive platform for precise neuromodulation in wild‐type animals.

We report a geometrically reconfigurable silk-based electronic implant designed for peripheral nerve monitoring. Metallic conductive structures were fabricated on the surface of a geometrically reconfigurable membrane using Micro-Electro-Mechanical Systems (MEMS) technology. By controlling the cutting angles of the membrane, we achieved water-responsive self-wrapping spiral electrodes, which exhibit excellent geometric adaptability to nerve tissues. In vitro tests demonstrate that our electrodes possess good mechanical, biocompatibility, and electrical properties. In vivo applications confirmed that our electrodes can effectively record compound nerve action potentials (CNAP). Our approach offers a new method for developing implantable electronic devices that require mechanical and geometrical adaptability.

We developed a flexible perforated microelectrode array to create a robust retina-electrode interface for light sensing and image recognition. The perforated structure of microelectrode array enables the application of gentle suction, securing the retina onto the electrodes for seamless contact. This design allows for the simultaneous recording of spike firings from hundreds of retinal ganglion cells in response to light stimuli. Our findings show that most cells respond actively to light, confirming the light-sensing capability of this interface. Integrated with machine learning algorithms, this interface supports accurate recognition of letters and colors based on spike firings.

Retina converts light stimuli into spike firings, encoding abundant visual information critical for both fundamental studies of the visual system and therapies for visual diseases. However, probing these spikes directly from the retina is hindered by limited recording channels, insufficient contact between the retina and electrodes, and short operational lifetimes. In this study, we developed a perforated and flexible microelectrode array to achieve a robust retina-probe interface, ensuring high-quality detection of spike firings from hundreds of neurons. Leveraging the retina's natural light-sensing ability, we created a hybrid bioelectronic system that enables image recognition through machine learning integration. We systematically explored the system's spatial resolution, and demonstrated its capability to recognize different colors and light intensities. Importantly, due to the perforated structure, the hybrid system maintained over 94 % accuracy in distinguishing light on/off conditions for 9 h ex vivo. Finally, inspired by the eye's configuration, we developed a bioelectronic mimic eye capable of recognizing objects in real environments. This work demonstrated that the hybrid bioelectronic retina-probe interface is effective not only for light sensing but also for efficient image and object recognition.

We present a novel modification for an ultra-flexible neural probe capable of simultaneous electrophysiological recording and dopamine detection within cerebral tissue. Following electrode modification, the electrochemical stability of the electrode coating was enhanced, and its specific surface area was increased, thereby significantly improving its sensitivity to dopamine. Building upon this foundation, the ultra-flexible design minimizes damage to brain tissue, enabling stable simultaneous detection of electrophysiological and electrochemical signals for over six weeks. The high-density electrode array design facilitates concurrent monitoring of activities across multiple brain regions. This novel approach provides new methodologies for neuroscience research and the treatment of brain disorders.

This paper reports a silk-based deep-brain neural interface featuring a deformable microelectrode array paired with a silk scaffold, designed to deployed and proper functioning in the dynamic milieu of the cerebrospinal fluid. The whole microelectrode array can be minimally invasively implanted into deep brain regions with the assistance of commonly used clinical catheters, self-unfold in the lateral ventricles to conformally attach to large-scale subcortical nuclei surfaces, and capture high-quality signals by virtue of the microclectrode's in-plane shielding. The results represent the inaugural demonstration of recording Parkinsonian electrophysiological biomarkers on the caudate surface in sheep, accompanied by high accuracy discrimination of pathological neural activity across multiple microelectrode sites.

High spatiotemporal resolution invasive neural signal acquisition provides significant benefits for neuroscience research and neurorehabilitation. However, full-rate acquisition of action potential (AP, also known as spike) results in substantial data bandwidth waste due to their inherent sparsity, which is unfavorable for low-power and miniaturized designs. This work presents a 128-channel system-on-chip (SoC) supporting both Full Bandwidth Recording (FBR) mode and Hybrid Recording (HR) mode. In FBR mode, the SoC performs neural recording at 20 kS/s sampling rate with 2.26 mu V-rms inputreferred noise (IRN) over a 1-10 kHz frequency range. In HR mode, the SoC directly digitizes spikes while recording local field potentials (LFPs) at 1 kS/s, reducing power consumption by 84% and data bandwidth by 94.7%. The SoC integrates impedance measurement, power management, clock generation, and digital control, enabling a high-density and compact design. Designed in SMIC 130 nm technology, the SoC consumes 11.8 mW in FB mode and 1.8 mW in SPD mode, with a die size of 4.45 mm x 4.45 mm.

This study introduces a novel flexible electrode designed as a self-rolling neural interface, aimed at advancing neurological diagnostics and modulation. The electrode is constructed from a prestressed metal and polymer multilayer structure that spontaneously assumes a controlled 3D geometry upon release from its substrate. By fine-tuning the fabrication process to adjust the residual mismatched stresses between layers, the electrode can be customized with varying radii of curvature to match the diameters of peripheral nerves in different animal models, such as SD rats and rabbits. The incorporation of mussel adhesive protein (MAP) ensures a stable attachment in liquid environments after curing. In vivo experiments have successfully demonstrated the electrode’s capability for selective neuromodulation and signal recording across 32 channels when encircling the sciatic nerve. The use of microcurrent programmable electrical stimulation near the recording channels yields high-quality electrophysiological signals. These findings highlight the potential of self-rolling flexible electrodes in modulating the sciatic nerve and capturing high-quality compound nerve action potentials (CNAPs).