Analysis Method for Regional Coupling Causal Relationships in EEG-fNIRS Bimodal Brain Data

Abstract

With the advancement of neuroscience research, single-modal EEG data has shown limitations in revealing complex brain activity mechanisms. While EEG (electroencephalogram) exhibits high temporal resolution, it lacks spatial localization accuracy, whereas fNIRS
(functional near-infrared spectroscopy) demonstrates superior spatial resolution but relatively insufficient temporal resolution. To overcome
these constraints, EEG-fNIRS multimodal data fusion technology has emerged as a research hotspot. Understanding brain region coupling
causal relationships remains pivotal for deciphering neural information processing mechanisms. This study focuses on methods for analyzing brain region coupling causal relationships in EEG-fNIRS multimodal data. First, we elucidate the characteristics of multimodal data and
fusion foundations, clarifying core features of single-modal data and the inherent logic of multimodal fusion. Subsequently, we explore the
fundamental concepts of brain region coupling and core principles of causal relationship analysis, constructing a logical framework for causal
inference in multimodal data. Building upon this foundation, we systematically review key technologies for causal relationship extraction and
propose targeted optimization strategies. Finally, practical implementation guidelines are summarized for major application scenarios, aiming to establish a scientific and systematic analytical methodology that provides robust support for elucidating neural mechanisms underlying
higher brain functions.