Distributed fMRI dynamics predict distinct EEG rhythms across sleep and wakefulness
The brain exhibits rich oscillatory dynamics that play critical roles in vigilance and cognition, such as the neural rhythms that define sleep. These rhythms continuously fluctuate, signaling major changes in vigilance, but the widespread brain dynamics underlying these oscillations are difficult to investigate. Using simultaneous EEG and fast fMRI in humans who fell asleep inside the scanner, we developed a machine learning approach to investigate which fMRI regions and networks predict fluctuations in neural rhythms. We demonstrated that the rise and fall of alpha (8-12 Hz) and delta (1-4 Hz) power-- two canonical EEG bands critically involved with cognition and vigilance--can be predicted from fMRI data in subjects that were not present in the training set. This approach also identified predictive information in individual brain regions across the cortex and subcortex. Finally, we developed an approach to identify shared and unique predictive information, and found that information about alpha rhythms was highly separable in two networks linked to arousal and visual systems. Conversely, delta rhythms were diffusely represented on a large spatial scale primarily across the cortex. These results demonstrate that EEG rhythms can be predicted from fMRI data, identify large-scale network patterns that underlie alpha and delta rhythms, and establish a novel framework for investigating multimodal brain dynamics. Author summaryNeurons often fire in synchrony, generating rhythms that play major roles in brain functioning. These rhythms are hallmarks of different brain states of vigilance, such as sleep and wakefulness. Sleep disorders are extremely prevalent among adults, and studying neural rhythms associated with vigilance states is a key step towards understanding sleep disorders and how healthy sleep can be restored. Measuring how neural rhythms affect the brain, however, is difficult: the primary method used in humans, electroencephalography (EEG), can only measure neural activity close to the scalp. EEG can be combined with functional magnetic resonance imaging (fMRI), which is capable of measuring activity in deep brain regions, but fMRI data can be difficult to analyze, as it estimates neural activity indirectly by measuring changes in blood oxygenation. We developed an approach to analyze combined EEG-fMRI data using machine learning, and used it to investigate how fluctuations in neural rhythms across sleep and wakefulness are tied to changes in neural activity throughout the whole brain. Our results describe how different brain networks are coupled to alpha and delta rhythms, and provide a new approach for analyzing EEG-fMRI data that can be employed to investigate other neural rhythms necessary for healthy brain functioning.