Gore, John C.; Li, Muwei; Schilling, Kurt G.; Xu, Lyuan; Li, Yikang; Zu, Zhongliang; Anderson, Adam W.; Ding, Zhaohua; & Gao, Yurui. (2026). The nature and interpretation of BOLD signals in white matter – A review. Magnetic Resonance Imaging, 127, 110596. https://doi.org/10.1016/j.mri.2025.110596
This review looks at recent research showing that blood oxygenation level–dependent (BOLD) signals in white matter (WM) contain meaningful information about brain activity. These signals are influenced by the structure of white matter, its blood supply, and its metabolism, and they are closely connected to functional MRI (fMRI) signals in gray matter (GM). BOLD signals in WM can be detected both during tasks and at rest, where their natural fluctuations reveal coordinated activity between white and gray matter. Even so, many fMRI studies have traditionally ignored WM signals or treated them as noise.
New evidence shows that WM BOLD signals reflect how different brain regions communicate. Studies have found that the strength and behavior of these signals depend on features such as myelination, neurite density, mitochondrial content, and blood vessels within white matter tracts. Different types of fibers, such as association and projection fibers, show different BOLD patterns, and some heavily myelinated fibers may show little or no detectable signal. Research has also clarified how WM BOLD signals relate to GM networks, including during resting-state activity. Together, these findings suggest that WM BOLD signals provide valuable insight into brain function and should be included in fMRI analyses to better understand how the brain is organized and operates.
Fig. 1. Population maps of HRF features show qualitative differences between GM and WM. Shown are the MNI T1, WM and GM masks for anatomical reference. Population-averaged features of the HRF are shown for FWHM, Height, PSC, time to Peak, Time to Dip, and Dip Height.