Abstract
Invasive neurointerfaces (Utah array, Neuropixels, Neuralink) demonstrate remarkable outcomes in short-term experiments; however, their durability and biocompatibility remain limited. The formation of glial scars, mechanical mismatch, and chronic inflammation reduce signal quality and constrain neuroplasticity. This article discusses the prospects of transitioning from rigid electrode-based systems to biological solutions, including autologous neural organoids and hybrid tissue interfaces, which are capable of integrating into the brain and supporting natural mechanisms of self-learning.
Keywords: Neurointerfaces, BCI, brain, biohybrid
Introduction
Neurointerfaces have become a key area of neuroscience and bioengineering, opening new prospects for restoring lost functions in patients with paralysis, neurodegenerative disorders, or sensory deficits. The first clinical platforms (BrainGate, Utah array, Neuralink) have confirmed the feasibility of directly recording and decoding neural activity to control external devices [Hochberg et al., 2012; Musk & Neuralink, 2019]. However, the issue of long term stability remains unresolved.
Limitations of Modern Invasive Interfaces
Contemporary invasive neurointerfaces (Utah array, Neuropixels, Neuralink) have demonstrated significant success in short-term experiments: patients with severe paralysis can control a cursor, type text, or interact with external devices directly through neural activity decoding [Hochberg et al., 2012; Musk & Neuralink, 2019]. However, the long-term stability of such interfaces remains a serious challenge.
Studies have shown that any implanted electrode is perceived by the brain as a foreign body. This triggers a cascade of immune responses, including microglial and astrocytic activation, glial scar formation, and encapsulation of the implantation site [Polikov et al., 2005; Campbell & Wu, 2018]. These processes are accompanied by a reduction in neuronal density near the electrode, an increase in impedance, and a decline in the signal-to-noise ratio. A key factor is the mechanical mismatch: soft brain tissue has a Young’s modulus on the order of tens of kilopascals, whereas rigid electrode materials reach hundreds of gigapascals [Gilletti & Muthuswamy, 2006]. As a result, natural micromotions of the brain (breathing, vascular pulsation) lead to chronic tissue trauma.
Clinical observations confirm these limitations. The Utah array can provide stable recordings for 1–2 years; however, the number of functional channels gradually decreases over time [Barrese et al., 2013; Prasad et al., 2014]. The new generation of devices, such as Neuropixels 2.0, demonstrates months of stable performance in animal models [Steinmetz et al., 2021], yet the prospects for long-term (>10 years) application in humans remain unclear. In 2024, Neuralink reported retraction of a portion of polymer threads in its first implanted patient, further highlighting the unresolved issue of chronic biocompatibility [Neuralink Clinical Update, 2024].
Impact on Neuroplasticity
The brain is not a static “wire” but a dynamic, self-learning network shaped by millions of years of evolution. Neuroplasticity is mediated through fine mechanisms of synaptic remodeling, including Hebbian plasticity and spike-timing dependent plasticity (STDP), where millisecond-scale correlations of activity and endogenous rhythms (theta, gamma, alpha) determine the strengthening or weakening of synaptic connections [Markram et al., 2012].
Invasive electrodes disrupt this harmony. First, the area surrounding the electrode becomes a “dead zone”: neurons degenerate, connections are severed, and local plasticity is diminished. Second, the very nature of electrical stimulation is coarse: instead of molecularly precise neurotransmitter signals, a relatively “noisy” current is delivered, simultaneously activating hundreds of cells. The effect may be noticeable in the short term (e.g., partial “restoration of vision” in visual prostheses), but in the long term, it does not sustain the physiological process of brain self-learning and, on the contrary, constrains it.
Experience with deep brain stimulation (DBS) in Parkinson’s disease illustrates this duality. On the one hand, patients achieve lasting relief of motor symptoms, indicating that stimulation can indeed induce plastic reorganization [Temel et al., 2006]. On the other hand, chronic stimulation may produce cognitive side effects— such as reduced cognitive flexibility, mood alterations, or even the emergence of pathological behavioral patterns [Krack et al., 2010]. This suggests that imposing an external rhythm interferes with the endogenous processes of brain self-regulation.
Thus, the problem with invasive interfaces lies not only in their limited survivability but also in their potential to suppress the deeper capacity for neuroplasticity. If the brain can teach itself through its intrinsic oscillations and rhythms, then a rigid implant delivering a “monotypic” signal is more likely to reduce this potential than to unlock it.
In response to these challenges, research efforts are increasingly shifting toward soft and biocompatible interfaces, as well as non-invasive modulation methods—such as transcranial magnetic and electrical stimulation (TMS, tDCS, tACS), ultrasound, optogenetics, and magnetoelectric nanoparticles. These technologies are aimed not at replacing the brain’s natural rhythms, but at their coordinated adjustment, thereby supporting—and in some cases enhancing—neuroplasticity.
Prospects of Biohybrid Interfaces
Given the fundamental limitations of current invasive neurointerfaces, it becomes evident that further progress is impossible without a paradigm shift. The problem lies in the mismatch between the scale and nature of implants and biological tissue. The brain is an organ shaped by millions of years of evolution, functioning at the level of individual cells, synapses, and molecular interactions. Its signaling dynamics are determined by millisecond-scale coincidences of activity, phase synchronization, and fine chemical modulation by neurotransmitters. Under such conditions, a massive electrode or a polymer thread transmitting a relatively “coarse” electrical signal essentially remains a mechanical crutch. The effect may be transient (e.g., visual prostheses or motor interfaces), but in the long run biology prevails: gliosis forms around the contact, neurons degenerate, and plasticity declines.
One potential solution is the transition from foreign materials to biologically harmonized interfaces based on autologous neural organoids. Organoids derived from induced pluripotent stem cells (iPSC) have demonstrated the ability to form functional neural networks with oscillatory activity comparable to that of the cortex [Trujillo et al., 2019; Sharf et al., 2022]. Upon transplantation, such structures integrate into the brains of animals: angiogenesis occurs, new synaptic connections are formed, and organoids begin to resonate with the endogenous rhythms of the host tissue.
Figure 1 Neural organoid in vitro. (A) Bright-field image showing the spherical structure of a neural organoid. (B) Transmission electron microscopy image illustrating cellular organization and peripheral layering. Scale bar = 10 μm.
The use of autologous cells eliminates the problem of immune rejection and allows the organoid to become part of the existing neural network. Unlike electrodes, which induce local neuronal death, an organoid may provide a novel substrate for neuroplasticity—additional resources for memory, learning, and network reorganization. Such an interface ceases to be a “mechanical adversary” and instead becomes a “biological ally”: it does not impose artificial patterns on the brain but expands its intrinsic repertoire.
Thus, the future of neurotechnology may lie not in the attempt to “wire the brain” but in the creation of biohybrid systems that constitute an extension of neural tissue itself. This marks a transition from a temporary and mechanical crutch to a technology that organically supports the brain’s natural rhythms, its capacity for development, and its ability to self-learn.
Conclusion
The brain is neither wiring nor a mechanical device but the outcome of millions of years of evolution—a biological system operating at the level of individual cells, molecules, and rhythms. Attempts to “connect” foreign constructs that have not been part of its evolutionary trajectory inevitably yield only short-term effects. In experimental settings, invasive neurointerfaces appear impressive: individuals deprived of movement or vision regain the ability to control a cursor, operate a prosthesis, or perceive stimuli again. Yet, the long-term prospects of such systems are limited: glial scarring develops at the site of contact, neurons either degenerate or withdraw from the electrode, signals weaken, and with them the potential for plasticity diminishes.
The thickness of a metallic or polymer conductor is incompatible with the scale of synaptic interactions, where impulse transmission occurs through molecular and even quantum-like mechanisms. Consequently, signals delivered by electrodes remain “coarse” relative to the fine dynamics of neural networks: they activate hundreds of cells simultaneously without reproducing the authentic language of the brain. This approach can be likened to a crutch: it compensates for a lost function but never becomes a natural extension of the organism.
Looking ahead, the future of neurointerfaces lies not in imposing coarse signals upon the brain but in creating solutions that align with its intrinsic rhythms and exploit its innate capacity for self-learning. A particularly promising direction is the development of biohybrid interfaces based on autologous neural organoids derived from a patient’s own cells. Unlike metal or silicon, such structures are capable of integrating into brain tissue, forming new connections, and expanding the substrate for neuroplasticity.
Thus, the central paradigm is shifting from the “mechanical adversary” to the “biological ally.” The neurointerface of the future is not a foreign wire but a biologically harmonized extension of the nervous system, one that sustains the brain’s natural rhythms and amplifies its capacity for growth and self-directed development.
References
- Brain micromotion around implants in the rodent somatosensory cortex
Aaron Gilletti, Jit Muthuswamy
https://pubmed.ncbi.nlm.nih.gov/16921202 - An Integrated Brain-Machine Interface Platform With Thousands of Channels
Elon Musk, Neuralink
https://pmc.ncbi.nlm.nih.gov/articles/PMC6914248 - Response of brain tissue to chronically implanted neural electrodes
Vadim S. Polikov, Patrick A. Tresco, William M. Reichert
https://pubmed.ncbi.nlm.nih.gov/16198003 - Spike-timing-dependent plasticity: a comprehensive overview
H. Markram, W. Gerstner, P. J. Sjöström
https://pubmed.ncbi.nlm.nih.gov/22807913 - Behavioural changes after bilateral subthalamic stimulation in advanced Parkinson disease: a systematic review
Yasin Temel
https://pubmed.ncbi.nlm.nih.gov/16621661 - Deep brain stimulation: from neurology to psychiatry
Paul Krack
https://pubmed.ncbi.nlm.nih.gov/20832128 - Functional neuronal circuitry and oscillatory dynamics in human brain organoids
Tal Sharf
https://www.nature.com/articles/s41467-022-32115-4
