Dr. Rothberg’s early entrepreneurial efforts, though integral to scientific progress, were shaky. He founded CuraGen, one of the first genomics companies, as a Yale graduate student in 1991. The company went public in 1999, and a year later had a market cap of $5 billion. It eventually crashed and was sold to small firm Celldex Therapeutics for $94.5 million in 2009.
While CEO at CuraGen, Dr. Rothberg’s newly born second child Noah was sent to the neonatal intensive care unit due to breathing issues. Noah ended up being fine, but Dr. Rothberg was frustrated there was no rapid test to ensure Noah didn’t have a congenital illness. He then founded CuraGen subsidiary 454 Life Sciences on the basis of massively parallel genome sequencing, which dropped the cost of sequencing 100-fold to $1 million. That was sold to Roche Diagnostics in 2007 for $140 million, and Roche closed down the company in 2013 after other sequencing approaches rendered 454’s technology noncompetitive.
He went on to invent semiconductor chip-based sequencing through his company Ion Torrent, founded in 2007. At the time, his firm’s Personal Genome Sequencing (PGM) DNA sequencer was the smallest and cheapest DNA decoder on the market, could read 10 million letters of genetic code in two hours, and sold for $50,000. His first true business success, he sold the company for $725 million in 2010. (Ion Torrent also paved the way for biotech firm Illumina Inc.’s famous “$1,000 Genome.” )
After Ion Torrent, Dr. Rothberg decided it was within his best interest to become a parallel entrepreneur instead of a serial one. In 2010, he heard a talk by MIT physicist Max Tegmark who was then working on connecting images from thousands of radio telescopes to measure distant stars’ energy. Developing what’s called a “butterfly network” that split data processing between all antennae was the secret to capturing quality images. Dr. Rothman approached Tegmark in 2011 with a proposition.
“I told Max that I’d love to use his ideas on imaging, and combine them with my recent success of DNA sequencing on semiconductors, to put an ultrasound machine on a chip,” Dr. Rothberg told MIT News. “I said I would start a company to do that…if he gave me the smartest students from MIT.”
He recruited Tegmark’s best student Nevada Sanchez and former Ion Torrent DNA-sequencing chip designer Keith Fife to found Butterfly Network. The startup aimed to make ultrasound imaging as simple and omnipresent as blood pressure or temperature checks, eventually bringing it to the home. With a number of MIT alumni joining the team in the next few years, Butterfly Network achieved FDA clearance for its Butterfly iQ iPhone-connected portable ultrasound scanner in 2017.
The secret: an easy-to-make semiconductor chip instead of the piezoelectric crystals used by traditional ultrasounds, allowing a mere $2,000 price tag for the system. The first users were doctors and clinicians familiar with ultrasounds, followed by paramedics, nurses, and doctors unfamiliar with the technology. There’s even a version for veterinary medicine. Eventually, Butterfly Network aims to sell directly to consumers.
Now Dr. Rothberg aims to make a more advanced imaging system portable—magnetic resonance imaging (MRI). His Hyperfine startup, part of the 4Catalyzer accelerator he founded in 2014, introduced the world’s first low-cost, point-of-care MRI system at the American College of Emergency Physicians (ACEP) Scientific Assembly during the last week of October.
“When MRI was launched,” Rothberg told STAT, “computing was 10 million times more expensive. Ten million times.” Dr. Rothman determined modern computing power could get a useable image from a much smaller magnet. Surely a bedside MRI device was within reach thanks to better computing power combined with advancements in magnets that had emerged from the green energy industry—those used in windmills, for example—to build one.
“We completely reexamined MRI from the ground up,” John Martin, M.D., Hyperfine’s chief medical officer, said in a press release. “Hyperfine’s system will open avenues to reach patients in clinical settings never before imagined for MRI.”
Hyperfine’s POC MRI creates standard clinical contrast images and informative 3D renders. The company is also developing software to improve each use through deep learning algorithms for image reconstruction and aiding in pathology diagnosis. According to Hyperfine, it’s 20 times less costly, 35 times lower power, and 10 times lighter than the most popular current MRI. It’s also currently pending FDA 510(k) clearance.
“Hyperfine is changing how medicine is practiced with point-of-care MRI,” Dr. Rothberg told the press. “More than just a revolutionary imaging system, Hyperfine is a platform for new applications. Our cloud-based software development kit will enable third parties to create AI applications for Hyperfine’s device. Ultimately, Hyperfine aims to facilitate stroke triage, continuous monitoring of unstable patients, and MRI guidance of surgical interventions.”