The California Institute of Technology has developed a new type of minimally invasive BMI for reading brain activity.

All tools for measuring brain activity have their drawbacks. Implanted electrodes (electrophysiology) can very accurately measure activity at the level of individual neurons, but require implantation in the brain.

Non-invasive techniques such as functional magnetic resonance imaging (fMRI) can show the entire brain, but require access to a large amount of equipment. Electroencephalography (EEG) does not require surgery, but can only measure activity at low spatial resolution.

Ultrasound works on the basis of high-frequency sound pulses. All measurements are based on sound vibrations that are reflected from body tissues. This method is commonly used to obtain images of the fetus in the womb and for other diagnostic information.

Ultrasound can also show the internal movement of organs. For example, red blood cells increase the pitch as they get closer to the source of the ultrasound waves and decrease as they move away. When scientists discovered this phenomenon, they were able to detect tiny changes in cerebral blood flow down to 100 micrometers (on the scale of the width of a human hair).

After the algorithm was trained, it was broadcast ultrasound data collected in real time from non-human primates.

As a result, the algorithm showed in a few seconds what the primate would do, for example, eye repetitions, movements, and turns.