There are more than 4 billion people globally who don’t have access to medical imaging—and could benefit from Butterfly iQ, a handheld ultrasound device. Jonathan Rothberg, a Yale genetics researcher and serial entrepreneur, figured out how to put ultrasound technology on a chip, so instead of a $100,000 machine in a hospital, it’s a $2,000 go-anywhere gadget that connects to an iPhone app. It went on sale last year to medical professionals. “Our goal is to sell to 150 countries that can pay for it. And [the Gates Foundation] is distributing it in 53 countries that can’t,” Rothberg says. The device isn’t as good as the big machines are and won’t replace them in prosperous parts of the world. But it could make scanning more routine. “There was a time when the thermometer was only used in a medical setting, when a blood-pressure cuff was only used in a medical center,” Rothberg says. “Democratizing [health] happens on multiple dimensions.”—Don Steinberg
Symptoms of lung cancer usually don’t appear until its later stages, when it’s difficult to treat. Early screening of high-risk populations with CT scans can reduce the risk of dying, but it comes with risks of its own. The U.S. National Institutes of Health found that 2.5% of patients who received CT scans later endured needlessly invasive treatments—-sometimes with fatal results—after radiologists erroneously diagnosed false positives. Shravya Shetty believes artificial intelligence may be the solution. Shetty is the research lead of a Google Health team that in the past two years built an AI system that outperforms human radiologists in diagnosing lung cancer. After being trained on more than 45,000 patient CT scans, Google’s algorithm detected 5% more cancer cases and had 11% fewer false positives than a control group of six human radiologists. The early results are promising, but “there’s a pretty big gap between where things are and where they could be,” says Shetty. “It’s that potential impact that keeps me going.”—Corinne Purtill