Real-time transcranial histotripsy treatment monitoring and localization using acoustic cavitation emission feedback
Background:
Transcranial MR-guided focused ultrasound (MRgFUs) has been explored as a non-invasive treatment option for various brain applications. To detect cavitation during these therapies, passive cavitation detection has been investigated, which requires an ultrasound transducer separate from the therapy transducer. Histotripsy is a cavitation-based ultrasound therapy which is being investigated for transcranial applications, and has been demonstrated to rapidly and accurately generate lesions through the skull. Cavitation events generated during histotripsy generate large acoustic cavitation emission (ACE) signals which can be detected through the skull using the elements of the histotripsy array as receivers. This study investigates the feasibility of using these ACE signals to monitor and localize cavitation activity in real-time during transcranial histotripsy treatments. The goal is to show that this is a viable technique for monitoring and localizing histotripsy-induced cavitation during treatments using the same transducer for both treatment and detection.
Material and Methods:
Histotripsy pulses were delivered through three excised human skullcaps using a 256-element, 500kHz, hemispherical histotripsy transducer with transmit-receive capable elements (Figure 1). Pulses were electronically steered through a 1cm diameter spherical volume centered at the geometric focus of the array through the skullcaps at PRF’s of 1Hz, 30Hz, and 100Hz to generate cavitation. During experiments, ACE signals were collected in real-time using the elements of the array as receivers. Treatment monitoring and localization was accomplished by back-projecting the acquired ACE signals into the field to form signal intensity maps of the transducer’s focal volume. To assess the effectiveness of the ACE-based cavitation monitoring and localization, concurrent optical images of the cavitation events were acquired using a multi-camera imaging system for comparison.
Figure 1-Experimental Setup
Results:
Using the ACE signals, we were able to monitor and localize cavitation activity in real-time through the skulls. There was good agreement between the locations of the bubbles found by back-projecting the ACE signals into the field and those measured using the cameras through all skullcaps, and deviations between the two measurements were generally bounded to within the measured diameters of the generated bubbles (<1.5mm, Figure 2, and Figure 3). The locations of the cavitation events predicted using ACE signals were preferentially biased in the prefocal direction of the array compared to camera measurements. The accuracy of the ACE predictions was observed to be affected by the PRF and deviated most from the camera measurements at 100Hz. This is likely due to the re-excitement of undissolved bubble nuclei in the field from previous pulses, which generate additional emission signals that act as noise in the ACE reconstructions.
Figure 2- ACE Localization vs. Hydrophone Measurements
Figure 3- ACE Localization vs. 3D Imaging
Conclusions:
In this study we have demonstrated the feasibility of using ACE signals to monitor transcranial histotripsy therapies in real-time. The locations of the bubble clouds predicted using the ACE signals, with respect to those measured via optical imaging, were found to be accurate to within the measured diameters of the generated bubbles. The performance of the ACE monitoring and mapping was found to be negatively impacted by higher PRFs, likely due to emissions from the re-excitement of undissolved bubble nuclei in the field from previous pulses. Strategies are being investigated to address this issue.