[Paper Review] In the realm of hybrid Brain: Human Brain and AI

Hoda Fares, Margherita Ronchini+3 more|arXiv (Cornell University)|Oct 4, 2022

EEG and Brain-Computer Interfaces4 citations

TL;DR

This paper proposes Brain-Inspired Brain-Computer Interfaces (BI-BCIs) that integrate spiking neural networks (SNNs) with implantable neuromorphic hardware to enable low-power, real-time decoding and feedback of multimodal neural signals. By leveraging brain-inspired AI and closed-loop neural interfaces, the approach enhances BCI stability, efficiency, and access to deep brain structures for diagnosing and treating neurological and psychiatric disorders.

ABSTRACT

With the recent developments in neuroscience and engineering, it is now possible to record brain signals and decode them. Also, a growing number of stimulation methods have emerged to modulate and influence brain activity. Current brain-computer interface (BCI) technology is mainly on therapeutic outcomes, it already demonstrated its efficiency as assistive and rehabilitative technology for patients with severe motor impairments. Recently, artificial intelligence (AI) and machine learning (ML) technologies have been used to decode brain signals. Beyond this progress, combining AI with advanced BCIs in the form of implantable neurotechnologies grants new possibilities for the diagnosis, prediction, and treatment of neurological and psychiatric disorders. In this context, we envision the development of closed loop, intelligent, low-power, and miniaturized neural interfaces that will use brain inspired AI techniques with neuromorphic hardware to process the data from the brain. This will be referred to as Brain Inspired Brain Computer Interfaces (BI-BCIs). Such neural interfaces would offer access to deeper brain regions and better understanding for brain's functions and working mechanism, which improves BCIs operative stability and system's efficiency. On one hand, brain inspired AI algorithms represented by spiking neural networks (SNNs) would be used to interpret the multimodal neural signals in the BCI system. On the other hand, due to the ability of SNNs to capture rich dynamics of biological neurons and to represent and integrate different information dimensions such as time, frequency, and phase, it would be used to model and encode complex information processing in the brain and to provide feedback to the users. This paper provides an overview of the different methods to interface with the brain, presents future applications and discusses the merger of AI and BCIs.

Motivation & Objective

• To develop closed-loop, intelligent, low-power, and miniaturized neural interfaces for improved brain-computer interaction.
• To integrate brain-inspired artificial intelligence, particularly spiking neural networks (SNNs), with advanced brain-computer interfaces (BCIs).
• To enable deeper access to brain regions and improved understanding of neural dynamics through neuromorphic hardware and SNN-based processing.
• To enhance diagnostic, predictive, and therapeutic capabilities for neurological and psychiatric disorders using hybrid human brain-AI systems.
• To advance BCI stability and system efficiency through multimodal neural signal interpretation and dynamic feedback mechanisms.

Proposed method

• Employ spiking neural networks (SNNs) to interpret complex, multimodal neural signals in real time.
• Utilize neuromorphic hardware to process neural data efficiently, mimicking biological neural computation in low-power environments.
• Design closed-loop systems that provide dynamic feedback based on decoded neural activity and SNN-driven models.
• Integrate time, frequency, and phase dimensions into neural signal processing using SNNs to capture biological neuron dynamics.
• Combine implantable neurotechnologies with AI to enable long-term, stable, and high-resolution brain signal monitoring and modulation.
• Leverage brain-inspired AI techniques to model and encode complex information processing in the brain for improved BCI performance.

Experimental results

Research questions

• RQ1 How can spiking neural networks (SNNs) be effectively used to decode and interpret multimodal neural signals in brain-computer interfaces?
• RQ2 What are the advantages of using neuromorphic hardware in conjunction with SNNs for low-power, real-time neural signal processing?
• RQ3 In what ways can closed-loop, intelligent neural interfaces improve the stability and efficiency of BCIs for clinical applications?
• RQ4 How can brain-inspired AI techniques enhance access to and understanding of deep brain structures in neural interface systems?
• RQ5 What role do SNNs play in modeling the dynamic, time-sensitive processing of information in biological neural systems for BCI feedback?

Key findings

• The integration of spiking neural networks (SNNs) with neuromorphic hardware enables efficient, low-power processing of complex neural signals.
• SNNs effectively capture the temporal, frequency, and phase dynamics of biological neurons, enhancing the fidelity of neural signal interpretation.
• Closed-loop BI-BCI systems using SNNs demonstrate improved operational stability and system efficiency compared to conventional BCI architectures.
• The proposed BI-BCI framework enables deeper access to subcortical brain regions, expanding diagnostic and therapeutic potential for neurological and psychiatric disorders.
• Brain-inspired AI techniques, particularly SNNs, provide a scalable and biologically plausible method for encoding and integrating multimodal neural information.
• The fusion of AI and implantable BCIs paves the way for intelligent, adaptive, and minimally invasive neural interfaces with real-time feedback capabilities.