To shed more light on this process, a French research team has developed a technique for detecting single ultrasound-induced nucleation events. Having demonstrated the technology in vitro, the researchers have now extended their work to detect bubble nucleation in living brain tissue (Phys. Med. Biol. 56 7001).

"The main motivation for our study was the lack of knowledge about the initiation of cavitation activity in biological tissue, compared to the behaviour of preformed or injected microbubbles," said Jérome Gateau of the Institut Langevin at ESPCI ParisTech. "The medical team of the Institut Langevin has previously investigated the possibility of using cavitation bubble signatures to guide transcranial ultrasonic treatments. We have also long been concerned by the safety of ultrasonic brain therapy, which is potentially affected by cavitation activity."

In vivo investigation

The researchers performed in vivo experiments on the brain tissue of eight anaesthetized sheep. Bubble nucleation was induced using a focused transducer with a central frequency of 660 kHz and peak negative pressures at focus ranging from –11.6 to –22.4 MPa. An average of 25 target positions per sheep were insonified, with excitations repeated to provide several attempts to nucleate in each region.

For the first six sheep, the detection sequence (25 successive exposures) was applied once per position, at a given amplitude. For the last two sheep, a sequence of 16 exposures was repeated more than once in each position, with the pressure amplitude gradually increased until bubble nucleation was detected.

The acoustically induced bubbles were detected using a linear array mounted on the side of the single-element transducer. The occurrence of bubble nucleation events was confirmed using successive passive recording and ultrafast active imaging.

"Passive detection provides information on the early phase of the nucleation event, while active detection provides information on the temporal behaviour of the induced bubbles," Gateau explained. "Combining the two improves the sensitivity of the technique by cross-confirming that a bubble nucleation event occurred."

Statistical study

The team then performed a statistical analysis of results from the 120 target positions to estimate the nucleation probability as a function of peak negative pressure. In vivo nucleation was found to begin at a peak negative pressure of –12.7 MPa, and was always induced when the peak negative pressure reached –22.4 MPa.

The researchers note that there was not a sharp transition in nucleation probability from almost never to almost always, but instead a slow increase was seen over a 10 MPa range. This confirms that bubble nucleation in vivo is a random phenomenon, and that defining a nucleation threshold at a given acoustic pressure may be over-simplistic.

For diagnostic ultrasound, the US Food and Drug Administration currently allows a mechanical index (a criterion used to evaluate the risk of inducing bubbles) of up to 1.9. At 660 kHz, this corresponds to approximately –1.5 MPa, which is about an order of magnitude lower than the acoustic pressure necessary to induce bubble nucleation in this work (–12.7 MPa at 660 kHz, which corresponds to a mechanical index of 15).

Another goal of this study was to evaluate the feasibility of using ultrasound pulses to induce bubble nucleation through the skull, for transcranial aberration correction. The researchers concluded that, accounting for the strong attenuation of the human skull (around –10 dB at 660 kHz), the high pressure amplitudes required to induce bubbles in the presence of only natural gas nuclei render this idea unlikely.

"This study showed that the specific risk of nucleating a bubble can be quantified in vivo, thereby contributing to a better evaluation of a safe use of medical ultrasound for brain imaging and brain therapy," Gateau told medicalphysicsweb. "Follow-up studies should result in an empirical criterion to assess this risk."

Gateau is currently refining the technique to perform a statistical study of the bubble nuclei population in blood. The medical team, meanwhile, is developing a clinical device for high-frequency (1 MHz) ultrasonic brain therapy that should limit bubble nucleation during brain insonation.

Related articles in PMB
In vivo bubble nucleation probability in sheep brain tissue
J Gateau et al Phys. Med. Biol. 56 7001
Optimal transcostal high-intensity focused ultrasound with combined real-time 3D movement tracking and correction
F Marquet et al Phys. Med. Biol. 56 7061
MR-guided transcranial brain HIFU in small animal models
B Larrat et al Phys. Med. Biol. 55 365
Non-invasive transcranial ultrasound therapy based on a 3D CT scan: protocol validation and in vitro results
F Marquet et al Phys. Med. Biol. 54 2597