Evaluation of auditory confounds in transcranial focused ultrasound brain stimulation using cortical recording

Abstract

In the past decade, transcranial focused ultrasound (tFUS) has emerged as a promising non-invasive brain stimulation technique, offering bimodal neuromodulation (i.e., excitation and inhibition) with exquisite spatial resolution and deep tissue penetration. In 2018, it was reported that tFUS-mediated neuromodulation may be accompanied by unintended auditory confounding effects, raising concerns about attributing observed response solely to the stimulation of targeted brain regions. Therefore, decoupling auditory confounds from tFUS-induced brain responses is crucial for accurately analyzing the effects of ultrasound neuromodulation. In this study, we aimed to perform electrophysiological assessments of auditory confounds associated with tFUS stimulation targeting non-auditory cortices. We recorded cortical responses via electroencephalography (EEG) or electrocorticography (ECoG) in urethane-anesthetized male Sprague-Dawley rats. The left primary auditory cortex (A1), along with bilateral motor, somatosensory, and visual cortices, was monitored during tFUS sessions. A custom-built 300- kHz miniature transducer was positioned over the right primary visual cortex. 0.2 s tFUS stimulations, pulsated with a 0.1 or 1 kHz pulse repetition frequency, were administered with a 5 s pulse train repetition interval at an acoustic pressure of 211 kPa or a spatial-peak temporal average intensity (ISPTA) of 750 mW/cm2 to be matched across different experimental conditions. In selected experiments, we also applied customized envelopes to the tFUS actuation signals to investigate the potential for reducing auditory responses. Cortical responses during tFUS stimulation were observed across all recorded regions, with characteristics suggestive of auditory confounding effects. When envelope shaping was applied, the amplitude of event-related potentials decreased, and normalized power in the delta to low-gamma frequency range (1-50 Hz) at A1 was attenuated up to 56%. This effect persisted even when the acoustic pressure was increased to match the ISPTA of conventional pulsed stimulation. Responses in other cortices, which exhibited temporal and spectral relationships with A1 activity, were also reduced when A1 responses were diminished through envelope shaping. Our findings demonstrate that the tFUSinduced auditory confounds are highly dependent on the envelope shape of the pulsed stimulation, rather than other sonication parameters. This study highlights the importance of waveform design in isolating the direct neuromodulatory effect of tFUS—an essential step towards its precise application in both research and clinical contexts.