This paper presents a probabilistic independent component analysis (PICA) approach for analyzing resting-state functional magnetic resonance imaging (fMRI) data. The method is optimized for fMRI data and is used to identify low-frequency resting-state patterns. The study demonstrates that PICA is effective in characterizing the spatio-temporal structure of resting-state data and that these patterns are consistent across subjects and resemble discrete cortical functional networks. The PICA approach is compared with seed-voxel-based correlation analysis, which is limited by the assumption that the seed voxel represents the entire network. PICA, on the other hand, does not require such assumptions and can identify overlapping spatial maps. The study also investigates the relationship between physiological noise and resting-state fluctuations, showing that PICA can separate physiological noise from resting-state patterns. The spatial characteristics of resting-state fluctuations are also examined, revealing that they are localized within grey matter and do not correspond to blood vessel networks. Finally, the consistency of resting-state patterns across subjects is investigated, showing that these patterns are localized within discrete areas of functional significance. The study concludes that PICA is a robust and effective tool for analyzing resting-state fMRI data and identifying low-frequency resting-state patterns.This paper presents a probabilistic independent component analysis (PICA) approach for analyzing resting-state functional magnetic resonance imaging (fMRI) data. The method is optimized for fMRI data and is used to identify low-frequency resting-state patterns. The study demonstrates that PICA is effective in characterizing the spatio-temporal structure of resting-state data and that these patterns are consistent across subjects and resemble discrete cortical functional networks. The PICA approach is compared with seed-voxel-based correlation analysis, which is limited by the assumption that the seed voxel represents the entire network. PICA, on the other hand, does not require such assumptions and can identify overlapping spatial maps. The study also investigates the relationship between physiological noise and resting-state fluctuations, showing that PICA can separate physiological noise from resting-state patterns. The spatial characteristics of resting-state fluctuations are also examined, revealing that they are localized within grey matter and do not correspond to blood vessel networks. Finally, the consistency of resting-state patterns across subjects is investigated, showing that these patterns are localized within discrete areas of functional significance. The study concludes that PICA is a robust and effective tool for analyzing resting-state fMRI data and identifying low-frequency resting-state patterns.