[Paper Review] Digital twin brain: a bridge between biological intelligence and artificial intelligence

Hui Xiong, Congying Chu+8 more|arXiv (Cornell University)|Aug 3, 2023

Molecular Communication and NanonetworksEngineering3 citations

TL;DR

This paper proposes the Digital Twin Brain (DTB) as a unified platform integrating brain atlases, multiscale neural models, and cross-modal simulations to bridge biological and artificial intelligence. By leveraging brain network organization and advanced computational modeling, DTB enables predictive, personalized simulations of brain function and disease, advancing artificial general intelligence and precision mental healthcare.

ABSTRACT

In recent years, advances in neuroscience and artificial intelligence have paved the way for unprecedented opportunities for understanding the complexity of the brain and its emulation by computational systems. Cutting-edge advancements in neuroscience research have revealed the intricate relationship between brain structure and function, while the success of artificial neural networks highlights the importance of network architecture. Now is the time to bring them together to better unravel how intelligence emerges from the brain's multiscale repositories. In this review, we propose the Digital Twin Brain (DTB) as a transformative platform that bridges the gap between biological and artificial intelligence. It consists of three core elements: the brain structure that is fundamental to the twinning process, bottom-layer models to generate brain functions, and its wide spectrum of applications. Crucially, brain atlases provide a vital constraint, preserving the brain's network organization within the DTB. Furthermore, we highlight open questions that invite joint efforts from interdisciplinary fields and emphasize the far-reaching implications of the DTB. The DTB can offer unprecedented insights into the emergence of intelligence and neurological disorders, which holds tremendous promise for advancing our understanding of both biological and artificial intelligence, and ultimately propelling the development of artificial general intelligence and facilitating precision mental healthcare.

Motivation & Objective

• To address the growing gap between biological intelligence and artificial intelligence by creating a unified computational framework.
• To develop a digital twin platform that integrates brain structure, function, and dynamics across multiple scales and modalities.
• To enable predictive modeling of brain function and neurological disorders using brain atlases as structural constraints.
• To support personalized virtual therapies by enabling individualized DTB construction through patient-specific data.
• To foster interdisciplinary collaboration toward artificial general intelligence and precision mental healthcare through a shared, open, and extensible simulation platform.

Proposed method

• Constructing the DTB using three core components: brain structure (from brain atlases), bottom-layer neural models for function generation, and a wide spectrum of applications.
• Leveraging brain atlases as foundational constraints to preserve the brain's network topology and organization in digital simulations.
• Integrating multiscale and multimodal data (e.g., neuroimaging, electrophysiology, genomics) into a unified modeling framework.
• Employing GPU-accelerated, numerically optimized simulation platforms to enable efficient, large-scale brain dynamics simulations.
• Applying pretraining and fine-tuning strategies on healthy control data followed by patient-specific adaptation for individualized DTB models.
• Designing a user-friendly, open-source platform with integrated modeling, simulation, visualization, and analysis capabilities to support researchers.

Experimental results

Research questions

• RQ1 How can brain atlases be used to constrain and structure digital twin models of the human brain?
• RQ2 What computational framework enables efficient, multiscale, and multimodal simulation of brain dynamics while preserving biological plausibility?
• RQ3 How can individualized digital twins be constructed to support personalized virtual therapies for neurological and psychiatric disorders?
• RQ4 What role does network architecture play in bridging the gap between biological intelligence and artificial intelligence?
• RQ5 How can digital twin brain models evolve from phenomenological to predictive models of brain disease?

Key findings

• The Digital Twin Brain (DTB) provides a transformative platform that unifies brain structure, function, and dynamics across multiple scales and modalities.
• Brain atlases serve as essential structural constraints that preserve the brain’s network topology, enabling biologically plausible modeling.
• Current neuroinformatic platforms lack full support for atlas-constrained, multiscale, and multimodal simulations, highlighting the need for a new, integrated platform.
• The DTB enables the transition from phenomenological to predictive modeling of brain diseases, supporting the development of personalized treatment strategies.
• Individualized DTBs can be built via pretraining on healthy controls followed by fine-tuning with patient-specific data, enabling virtual therapy simulations.
• The DTB holds significant promise for advancing artificial general intelligence and precision mental healthcare by simulating the emergence of intelligence and neurological dysfunction.