Using independent component analysis to extract a cross-modality and individual-specific brain baseline pattern

•MIC exhibits a high intra-subject stability across modalities and over months.•Identification using MIC is more accurate than functional connectome fingerprinting.•Stability of MIC is accompanied by high explained variance.•Noise and scan duration have limited impact on the extraction of MIC.•MIC could be considered as a baseline pattern of the ongoing brain activity. The ongoing brain activity serves as a baseline that supports both internal and external cognitive processes. However, its precise nature remains unclear. Considering that people display various patterns of brain activity even when engaging in the same task, it is reasonable to believe that individuals possess their unique brain baseline pattern. Using spatial independent component analysis on a large sample of fMRI data from the Human Connectome Project (HCP), we found an individual-specific component which can be consistently extracted from either resting-state or different task states and is reliable over months. Compared to functional connectome fingerprinting, it is much more stable across different fMRI modalities. Its stability is closely related to high explained variance and is minimally influenced by factors such as noise, scan duration, and scan interval. We propose that this component underlying the ongoing activity represents an individual-specific baseline pattern of brain activity.