DeepBrain: A novel human brain interface that non-invasively writes using focused ultrasound

How can we build direct interfaces with the human brain? Current approaches are looking at implantable electrodes. However, due to the risks of surgery, while implants are fine for patients, they are not appealing to the general population. Even patients are often reluctant: despite increasing numbers of cochlear implant surgeries, uptake in the adult patient population is only 5.6%.

This proposal aims to test a novel non-invasive neural interface that uses ultrasound to directly stimulate (write onto) the human brain. Ultrasound can have a high spatial resolution, reach deep-brain structures, and can be used to both elicit neural activity (focused ultrasound neuromodulation) and to read neural activity (functional ultrasound imaging). Unfortunately, the barrier to realising the potential of this technology is that the skull is in the way of sound propagation, absorbing most of the acoustic energy. This makes higher frequency ultrasound with lower wavelengths and thus a smaller focus of stimulation unfeasible. As a result, current deep-brain foci are more than 10mm large preventing the precise stimulation of targets within a region.

To realize high-resolution stimulation, we will avoid sound propagation through the skull and test a potential alternative route. This approach will allow 'writing' using higher frequencies (>1 MHz) compared to the currently used frequencies of 0.5 MHz for transcranial stimulation. Due to the higher frequency, we can therefore write at higher spatial resolutions.