Modular Brain-Machine Interface: Innovative Breakthrough in Neurorecording, Neurostimulation, and Drug Delivery

Academic Background

Brain-Machine Interfaces (BMIs) are crucial tools in neuroscience and clinical medicine, enabling interactions between the brain and the external world through charge, matter, and information exchange. They are widely used in neural decoding, diagnosis and treatment of neurological diseases, and brain science research. With the advancement of neuroscience, multimodal BMIs have garnered significant attention for their ability to support functions such as neurorecording, neurostimulation, and drug delivery. However, most existing multimodal BMIs are designed for specific scenarios with highly integrated fixed configurations, making it difficult to adapt to different experimental needs.

To address this issue, Sheng et al. proposed a modular multimodal BMI, aiming to provide flexible modular design that allows the interface to adjust configurations, modalities, and functions according to varying experimental requirements. This design not only enhances the adaptability of the device but also offers a universal platform for experiments requiring multiple modalities and specifications.

Source of the Paper

This paper was co-authored by Tiancheng Sheng, Lingyi Zheng, Jingwei Li, and others, with the team hailing from the School of Biomedical Engineering, Department of Mechanical Engineering at Tsinghua University, and the State Key Laboratory of Robotics at the Shenyang Institute of Automation, Chinese Academy of Sciences. The paper was published in Device on May 16, 2025, titled Modular Brain-Machine Interface for Neurorecording, Neurostimulation, and Drug Delivery.

Research Process and Results

1. Design and Development of the Modular BMI

The research team proposed a modular wireless multimodal BMI based on the principle of functional decentralization, decomposing complex systems into multiple simple modules. The device consists of a Support Layer (SL), a Functional Layer (FL), and an Interface Layer (IL). The SL includes functions such as wireless communication, power management, and system management, while the FL comprises modules for neurorecording, neurostimulation, and drug delivery. The IL directly interacts with the experimental subjects.

To simplify connections between modules, the team designed a unified physical interface (Zero Insertion Force connector, ZIF connector) and communication protocols (Serial Peripheral Interface, SPI, and Pulse Width Modulation, PWM), allowing different functional modules to be plug-and-play. This modular design not only improves the device’s flexibility but also provides the potential for future expansion with other functional modules (e.g., optical modules, temperature sensors, etc.).

In terms of hardware design, the SL adopts an ARM-based Microcontroller Unit (MCU) and Wi-Fi transceiver circuits, while the FL includes a 64-channel neurorecording module, a 16-channel neurostimulation module, and a micropump-based drug delivery module. The driver codes for all modules are integrated into the MCU, enabling users to monitor signals and control various neuromodulation parameters in real time through dedicated software.

2. Experimental Validation and Application Scenarios

To validate the applicability of the modular BMI, the research team conducted experiments in four different scenarios:

Scenario 1: Closed-loop Seizure Modulation in Freely Moving Rats

The team validated the BMI’s capability in closed-loop drug delivery using a rat model. In the experiment, epileptic activities were recorded using a foldable ECoG electrode implanted in the left parietal lobe, while a microtube was implanted in the right hippocampus for drug delivery. Epileptiform activities were induced using 4-Aminopyridine (4-AP), and upon real-time detection of seizure events, GABA (γ-aminobutyric acid) was immediately injected for modulation. The results showed that GABA delivery significantly suppressed epileptic activities, demonstrating the device’s effectiveness in closed-loop neuromodulation.

Scenario 2: Multi-channel Neurorecording in Swine

In an experiment involving acute cortical recording in swine, the team used a 128-channel ECoG electrode, validating the device’s applicability in large animal models. Neural activities were recorded from sites of varying diameters and monitored in real time using computer software. The results showed that microrecording sites captured signals similar to those from larger areas, highlighting the device’s potential in cortical recording.

Scenario 3: EEG Recording from Human Scalps

The team developed an ear-mounted configuration using a 32-channel flexible EEG patch electrode to record EEG signals from human scalps. The results demonstrated the device’s ability to detect differences in alpha waves between eyes-closed and eyes-open states, confirming its application value in human EEG recording.

Scenario 4: In Vitro Directional Neurostimulation

To validate the device’s capability in Directional Deep Brain Stimulation (DDBS), the team designed an omnidirectional electrode based on a Micro-Electro-Mechanical System (MEMS). The experiment showed that the electrode could generate three-dimensional electric fields by adjusting stimulation currents, demonstrating its potential in deep brain stimulation.

3. Experimental Conclusions and Value

The modular BMI developed by the research team successfully achieved seamless switching of multiple functions through unified interfaces and flexible modular design, providing a highly adaptable platform for neuroscience research. The device’s value lies not only in its multifunctionality but also in its broad applicability across different experimental scenarios, such as seizure modulation, cortical recording, EEG detection, and deep brain stimulation.

Moreover, the modular design provides a foundation for future expansion with other functional modules (e.g., optical modules, temperature sensors, etc.), further enhancing the device’s potential. The research team validated the effectiveness of each function through experiments, offering reliable evidence for future clinical and research applications.

Research Highlights

1. Modular Design: Adopts a unified physical interface and communication protocols to enable plug-and-play functionality, enhancing the device’s flexibility and scalability.
   
2. Multifunctionality: Supports multiple functions, including neurorecording, neurostimulation, and drug delivery, making it suitable for various experimental scenarios.
   
3. Broad Applications: Demonstrates excellent performance in both animal models and human experiments, validating its applicability across different research needs.
   
4. Innovative Experimental Methods: Utilizes new technologies such as foldable ECoG electrodes and omnidirectional electrodes, providing new tools for neuroscience research.

Additional Valuable Information

The research team also developed dedicated user software to support real-time signal monitoring and parameter adjustment, further improving the device’s usability. Additionally, the paper details the design and fabrication processes of the electrodes, providing valuable references for researchers in related fields.

This innovative modular BMI offers a powerful tool for neuroscience research and is expected to drive further advancements in neuroscience, clinical medicine, and brain-machine interface technology.