Adaptive Transcranial Ultrasound Doppler Imaging of the Brain

The development of fully noninvasive, transcranial functional ultrasound (fUS) would increase the translatability and clinical potential of this neuroimaging modality. Unfortunately, transcranial fUS is hindered by skull-induced aberrations which degrade the power Doppler image quality and lower sensitivity. As a result, a majority of fUS imaging studies rely on craniotomies or acoustically transparent cranial windows. To advance fUS technology further, we present an adaptive aberration correction method based on ray-tracing through four tissue layers (transducer lens, gel and skin, bone, and brain tissue). Our method segments these layers and estimates ultrasound wave speeds in each layer iteratively. Once a velocity model of the imaging plane of interest is retrieved, ultrafast power Doppler imaging of the brain is performed using a ray-tracing beamformer that accounts for wave refraction. We tested our method in three adult rats, and estimated wave speeds for the skin/gel layer ( $1628 \pm 7 \mathrm{~m} / \mathrm{s}$ ), skull bone ( $3247 \pm 110 \mathrm{~m} / \mathrm{s}$ ), and brain tissue ( $1526 \pm 55 \mathrm{~m} / \mathrm{s}$ ). After aberration correction, we measured an average adult rat skull thickness of $388 \pm 41 \mu \mathrm{~m}$ in agreement with anatomical records. The largest improvements in the Doppler imaging quality were observed in cortical brain layers adjacent to the skull; specifically, lateral spatial resolution was improved by 32 %. Our method consistently outperformed Doppler imaging based on traditional delay-and-sum (DAS) beamforming, which assumes a uniform sound speed.