Fully implantable wireless brain-computer interface for humans: Advancing toward the future

On January 29, 2024, Elon Musk made a public announcement via social media about the successful implantation of a Neuralink device in humans. Shortly after, a collaborative team from Xuanwu Hospital and Tsinghua University in China revealed the advancements in their clinical experiment testing the Neural Electronic Opportunity (NEO)—a wireless brain-computer interface (BCI).1Deng C. Li X. Dai J. Challenges for translating implantable brain-computer interface to medical device.Innovat. Med. 2023; 1100040https://doi.org/10.59717/j.xinn-med.2023.100040Crossref Google Scholar These simultaneous developments sparked widespread interest and renewed enthusiasm for BCIs worldwide. The concept behind BCI is to establish a direct communication pathway between the brain and an external device. As a result, an integrated BCI must have the capability to record, interpret, and stimulate neural activity, as well as interact with external devices. Over the past few decades, due to technical challenges, most BCI advancements have been confined to animal experiments or human subjects with wired connections, thus impeding their use in treating neurological disorders or augmenting human abilities. The recent breakthrough achieved by Neuralink marks the first instance of successfully implanting a completely wireless BCI device in the human brain and effectively detecting spiking activity. Notably, the device, known as N1, is only the size of a coin yet is fully equipped to record and transmit brain signals, recharge wirelessly, and communicate via Bluetooth (Figure 1). Similarly, the NEO is also a wireless device, albeit slightly larger (about the size of two parallel coins). The NEO is powered by near-field communication (NFC) and communicates via NFC as well, making it battery free. The primary distinction between these two devices is that N1 directly interfaces with the brain cortex, while NEO is positioned above the dura (Figure 1). These two approaches, namely full-invasive and semi-invasive, represent two potential future paths for BCI development. Full-invasive BCIs provide precise signal sources for decoding but inevitably entail intrusion into the intact brain. Semi-invasive BCIs, in theory, do not harm the cortex but can only gather signals from the vicinity of the cortex. However, the results of using N1 in humans are currently unknown. In contrast, the NEO has been successfully implanted in a patient with a spinal cord injury for over 3 months. This patient was able to control an external device to drink water after undergoing rehabilitation training. It is worth noting that decoding motion intention does not necessarily require a spike-level signal or high-throughput recording. These two clinical trials collectively illustrate the engineering pathway to achieve fully wireless and implantable BCIs in humans, a crucial development for broadening their potential applications in the future. However, their broader implications in scientific or practical contexts are quite limited. The initial success of N1 has yet to prove its efficacy in signal collection, communication with external devices, and clinical outcomes, nor has it provided any information on data security, biocompatibility, and longevity. As for NEO, despite its notable success in aiding patients with spinal cord injuries, several issues persist. Firstly, while claiming to function throughout one's lifetime without a battery, NEO has not yet demonstrated its longevity and long-term biocompatibility. Secondly, due to its semi-invasive approach, NEO can only capture signals from superficial layers near the dura, potentially impacting its decoding performance and limiting its expansion to other clinical applications. Thirdly, NEO is powered by the NFC technique, which relies on an extracorporeal machine using electric power, raising concerns about portability and resistance to interference. For instance, it would be impractical to carry a heavy external power source to ensure the device's functionality in future scenarios, and the device should consistently operate without interference in complex environments. However, these two experiments represent a significant advancement for BCIs and inspire greater possibilities for their future. Prior to transitioning to commercial or widespread medical use, additional efforts are required to address the technical challenges mentioned, confirm biosafety and efficacy, and overcome policy barriers. Nonetheless, the current state of BCI technology still falls short of the expectations of most individuals. For instance, individuals may have envisioned utilizing a BCI to upload and download memories, potentially saving significant time spent on reading, learning, and memorizing. However, the obstacle to realizing this vision lies not in the BCI hardware but rather in our lack of understanding of how memories are stored and retrieved in the brain. Comparable challenges extend to other potential application scenarios, such as transferring an individual's consciousness to digital form to achieve immortality, treating mental illnesses or neurodegenerative diseases, and restoring perception and cognitive functions. All of these endeavors depend on a deeper comprehension of the brain's operational principles.2Shan L. Huang H. Zhang Z. et al.Mapping the emergence of visual consciousness in the human brain via brain-wide intracranial electrophysiology.Innovation. 2022; 3100243https://doi.org/10.1016/j.xinn.2022.100243Abstract Full Text Full Text PDF Scopus (3) Google Scholar At the present stage, more achievable objectives for BCI could involve applications in motion-related function rehabilitation, recovery, or replacement, such as assisting paralyzed patients with regaining the fundamental ability to live like able-bodied individuals. In comparison to other functions, attaining motor control is relatively more feasible. With the aid of external executing devices, the BCI only needs to accurately interpret the brain's motion intentions.3Willett F.R. Avansino D.T. Hochberg L.R. et al.High-performance brain-to-text communication via handwriting.Nature. 2021; 593: 249-254https://doi.org/10.1038/s41586-021-03506-2Crossref PubMed Scopus (354) Google Scholar At the same time, it is crucial to set the industry standard for BCI technology in order to guarantee smooth interaction among various devices from different manufacturers, which is essential for the widespread use and spread of this technology. To achieve this, device manufacturers, academic institutions, and relevant industry associations need to come together and negotiate to reach a consensus. In the current phase, it is advisable to temper expectations regarding Elon Musk's endorsement of these initial clinical trials. As a businessman and the proprietor of Neuralink, Elon Musk has his own vested interests and requirements. He is adept at leveraging his influence to garner attention and potential investments through exaggeration and grandiose displays. However, he has not addressed the potential risks nor outlined strategies to mitigate or minimize them. In our view, the most significant risk is that the advancement of BCI may one day surpass human control. With artificial intelligence (AI) also rapidly progressing,4Huang T. Xu H. Wang H. et al.Artificial intelligence for medicine: Progress, challenges, and perspectives.Innovat. Med. 2023; 1100030https://doi.org/10.59717/j.xinn-med.2023.100030Crossref Google Scholar,5Xu Y. Liu X. Cao X. et al.Artificial intelligence: A powerful paradigm for scientific research.Innovation. 2021; 2100179https://doi.org/10.1016/j.xinn.2021.100179Abstract Full Text Full Text PDF Scopus (314) Google Scholar BCI offers a conduit for AI to interface with the human mind, potentially paving the way for AI to exert control over humans. Therefore, the development of BCI (and AI) is a double-edged sword. While advanced BCI technology undoubtedly holds benefits for human well-being, it is crucial to objectively assess risks and formulate strategies and policies to prevent or mitigate them. Public opinion should not be swayed by a small cohort of individuals who wield cutting-edge technology. This work was supported by the National Natural Science Foundation of China (32371066), the Guangdong Basic and Applied Basic Research Foundation (2022A1515010134), the Shenzhen-Hong Kong Institute of Brain Science–Shenzhen Fundamental Research Institutions (NYKFKT2019009), and the Shenzhen Technological Research Center for Primate Translational Medicine (F-2021-Z99-504979). The authors declare no competing interests.