Chronic functional ultrasound imaging

of the rat brain through a thinned skull

The use of ultrasound to visualize blood vessels and monitor blood flow is a common procedure in biology and medicine. Unfortunatly, ultrasound imaging could not be used in brain as the thickness of the skull introduces a strong attenuation of ultrasound.

We developed functional ultrasound (fUS) using planar ultrasonic waves rather than focused beams. This method was applied with success to the measure of cerebral blood volume (CBV) in the brain at better spatiotemporal resolution than with other functional brain imaging modalities. Nevertheless, this technique was initially performed in rats after an invasive craniotomy restricting fUS imaging to acute experiments only. Here, we present a new strategy for chronic in vivo fUS imaging.


We report that after thinning the skull bone on a 1cm2 window (AP= ß+4 mm to -6 mm, L = +/- 5 mm above the midline), CBV could be visualized not only in cortex but also in deeper subcortical structures including thalamus and hypothalamus. The thinned-skull did not cause brain inflammation or neuronal death, indeed immunostaining for GFAP (microglial marker), Iba1 (astrocytic maker) and NeuN (neuronal marker) was not modified one day or three weeks after surgery.


Morphology of the vascular system was followed in a 3 weeks longitudinal study on 6 Sprague Dawley rats (250 g) anesthetized with either isoflurane or ketamine. No changes in the brain vascular system were noticed between the first and fifth fUS imaging session. Brain activation was elicited by electrical stimulation of hindpaw, forepaw or whisker pad (200 µs pulses, 5 Hz, 4s duration, 1 mA).



A. Immunohistochemalstudy of inflammatory response after fronto-parietal bone thinning. Left panel: GFAP, Iba1 and NeuN expression were not modified after 3 days or after 3 weeks (data not shown) after surgery.

B. Power Doppler images of the brain vascular system in chronic configuration 15 days after surgery.

CBV changes were followed at high spatial (80 µm) and temporal (0,5s) resolutions. With a single trial (no averaging), CBV increase was observed between 0,5s and 1s in the homotypic contralateral somatosensory cortex. CBV was also increased in the thalamus. We noticed a decrease in CBV in the ipsilateral somatosensory cortex with ketamine anesthesia but not with isoflurane. Decreased BOLD signals in the ipsilateral somatosensory cortex after activation were also previously observed by others (Boorman et al., J Neurosci 30, 4285-4294, 2010; Schäfer et al., Neuroimage 59, 3119-3127, 2012). Current development of a miniaturized ultrasonic probe should open the use of fUS on freely-moving animals.

A. Correlation map obtained 1, 7, 15 or 30 days after surgery in the same animal (electrical stimulation of hind paw, 200 µs pulses, 5 Hz, 4s duration, 1 mA.

B. Left panel: Blood volume changes in the activated area followed by fUS. Center panel: Detail of one trial showing the shape of the hemodynamic response. Right panel: Extracellular electrophysiological recording in the same area.