Editorial: Generative AI for brain imaging and brain network construction

Brain imaging plays an important role in exploring brain which is claimed to be the most complex thing in the universe. In recent years, there has been remarkable progress in the field of brain imaging techniques. Meanwhile, advancements in brain imaging technologies have greatly enhanced our understanding of brain networks. As an important tool for exploring the relationship between brain structure and function, brain network computation represents the most promising direction for artificial intelligence to achieve breakthroughs and advancements in the field of neuroscience. By employing artificial intelligence algorithms to integrate complementary features from multi-modal brain imaging, it is possible to uncover the connectivity characteristics of neural circuits and establish multi-level mapping brain network models based on function, structure, and organization.

Generative artificial intelligence (AI) has witnessed significant expansion, encompassing the utilization of available data to generate fresh content that exhibits comparable underlying patterns to real-world data. The fusion of these two realms, generative AI and neuroimaging, offers a promising path for delving into diverse domains of brain imaging and brain network computation, specifically in the realms of extracting spatio-temporal brain characteristics and reconstructing the topological connectivity of brain networks. Generative artificial intelligence assists researchers in learning and understanding brain functional mechanisms in a broader feature space under limited sample conditions. It aids researchers in designing efficient fusion methods capable of handling and correlating multimodal data and domain knowledge information. By integrating multimodal brain data with prior knowledge from neuroscience, it achieves complementary synergy between the cooperative semantic information and knowledge rules inherent in different levels and factors of the brain.