Null Subtraction Imaging for functional ultrasound brain activation mapping

Null Subtraction Imaging (NSI) is a nonlinear beamforming approach that combines multiple receive apodizations and subtraction to enhance spatial resolution in ultrasound imaging. In NSI, a DC offset parameter is introduced in the apodization design to control the sharpening of the effective beam pattern, thereby tuning the spatial resolution. Here, we apply NSI to functional ultrasound (fUS) imaging of the brain and compare its performance with that of conventional delay-and-sum (DAS) beamforming. To this end, we conducted numerical simulations of cerebral blood volume (CBV) variations and experiments on anesthetized wild-type mice subjected to periodic vibrissae stimulation. Activation maps were computed by correlating the simulated or acquired CBV signals with the stimulation pattern. Several complementary metrics were used to characterize spatial localization, boundary sharpness, vascular overlap, signal-to-noise ratio (SNR), and functional detectability. Numerical simulations allowed us to distinguish improvements in spatial resolution from SNR-driven changes in the activation maps. Simulation results show more spatially localized activation maps with NSI and demonstrate that this improvement in spatial resolution is independent of detection sensitivity, although the resolution advantage of NSI depends on the input SNR of the imaging system. Experimental results are consistent with the simulations, demonstrating more spatially localized activation maps with NSI. At low DC offsets, however, NSI exhibits increased CBV signal noise and reduced detection sensitivity relative to DAS, reflecting a trade-off between spatial localization and sensitivity. An operating range around DC = 0.5 provides a practical balance between these effects in our imaging configuration. Together, these results establish NSI as a promising strategy for enhancing spatial localization in functional activation mapping with fUS imaging and provide a quantitative framework for guiding DC parameter selection.