The scanner being developed by entrepreneur Jonathan Rothberg’s startup company, Butterfly Network, doesn’t actually use X-rays or allow people to see through walls, but it does allow doctors to see real-time, 3-D images of a patient’s internal organs using ultrasound (high-frequency sound waves). Traditional ultrasound tests are usually performed by using a transducer that transmits sound waves through a gel into the body, forming an image from the waves that reflect off of the target tissues. According to one of Butterfly Network’s patents, Rothberg wants the device to help doctors become more effective in their diagnoses by allowing them to see “high resolution, volumetric three-dimensional images in a short time.”
Current medical imaging technologies typically have to compromise speed or the resolution of the resulting images for the sake of the other. Ultrasound tests, for example, though relatively quick and inexpensive, capture sonograms that pale in comparison to a high resolution MRI scan. However, the lengthy acquisition period and high cost of MRI scans render them impractical in emergency situations.
Image Source: Digital Vision
Rothberg has yet to disclose the device’s specifications or what the device will look like, but he has stated that it will be small, function as a phone accessory, and be in the price range of a few hundred dollars. He’s also described the medical imaging process: users will follow on-screen instructions on their phone similar to Apple’s Panorama snapshot tool, the device will capture a live, 3-D image and then send it to a cloud service, where various analysis programs will help automate diagnoses by highlighting key characteristics in the image. It has been designed to function much like X-ray vision does; it effectively opens a window into the patient’s body for the user to look through. The device will also utilize a form of artificial intelligence called “deep learning,” which is used by tech giants Google, Facebook, and Microsoft to improve the accuracy of its image analysis — the more images it captures, the smarter it gets at analyzing them and recognizing patterns.
Ultrasound chips like the ones being used in the Butterfly Network’s device have been around since 1994, but no commercial devices have ever been successful. Rothberg plans to have the first of the devices available in 18 months, with the goal of making it cheap enough for use around the world. The inherent intelligence of the device from “deep learning” will make it easy for any practitioner to use it, without training.
The Butterfly Network is only the first company to come out of Rothberg’s startup incubator, 4Combinator, which aims to transform “medical imaging, pharmaceutical development, and diagnostics,” according to their website.
