If you’ve welcomed a new baby into the world in the past few years, chances are you’ve also had a brush with 3D ultrasound: a noninvasive technique that uses sound to image soft tissue, internal organs, and—yes—even babies in three dimensions. But this relatively new technology has its drawbacks. Because ultrasound relies on tiny, low-volume echoes, most 3D imaging devices require thousands of sensors to generate a detailed picture. Now, engineers have developed a new system that cuts this number down to just one, they report today in Science Advances. They did so using compressive sensing, a technique that extrapolates information from one data source into many different points using some pretty complicated math. Engineers place an irregularly shaped piece of plastic, with small dots of varying heights on the surface, over the sensor. When the high-frequency sound waves emitted by the sensor pass through the plastic, this so-called “coded aperture mask” introduces small delays into the audio. When the sound bounces back to the sensor, it is then run through a computer algorithm that picks out the delays and represents them as individual pixels. Because the dots are randomly distributed across the surface of the cover, rotating the mask allows more measurements to be performed, and helps resolve the image in higher detail. The researchers tested their device by submerging two 3D-printed plastic letters in a small tank of water. When they activated the ultrasound sensor, it was able to pick the letters out of the data and pinpoint their location in 3D space.
Click here for free access to our latest coronavirus/COVID-19 research, commentary, and news.
Support nonprofit science journalism
Science’s extensive COVID-19 coverage is free to all readers. To support our nonprofit science journalism, please make a tax-deductible gift today.