From blood oxygenation level dependent (BOLD) signals to brain temperature maps

Area of Science:

• Neuroimaging
• Biophysics
• Computational Neuroscience

Background:

• Functional brain activity involves complex changes in cerebral blood flow (CBF) and oxygen consumption (CMRO₂).
• These physiological changes are known to induce local temperature variations within the brain.
• The Blood Oxygenation Level Dependent (BOLD) signal is a widely used indirect measure of neural activity, but its relationship with temperature is not fully understood.

Purpose of the Study:

• To develop a theoretical framework for converting BOLD images into quantitative brain temperature maps.
• To investigate the relationship between BOLD signals, CBF, CMRO₂, and brain temperature changes.
• To compare the spatial distribution of temperature changes with BOLD activations in a functional experiment.

Main Methods:

• Developed a theoretical model linking CBF and CMRO₂ changes to the BOLD signal using an oxygen limitation model.
• Derived a transcendental equation solved via the Lambert W function to relate physiological changes to the BOLD signal.
• Integrated this into the dynamic bioheat equation, solved numerically for each voxel to generate temperature maps.
• Applied the method to BOLD data from an attention to visual motion experiment.

Main Results:

• Generated brain temperature maps with variations in the range of (-0.15, 0.1) °C, consistent with experimental findings.
• Observed similar spatial distribution patterns between temperature and BOLD activations.
• Identified significant temperature activation in the precuneus, a region not detected by BOLD maps.
• Demonstrated that temperature maps exhibit greater localization to gray matter compared to BOLD maps.

Conclusions:

• The proposed framework enables the conversion of BOLD data into brain temperature maps.
• Brain temperature mapping provides complementary information to BOLD, highlighting specific regions like the precuneus.
• Temperature maps offer improved spatial specificity, concentrating activations in gray matter, thus enhancing the resolution of functional neuroimaging.