Pages

Saturday, November 10, 2018

Alums from this failed health tech start-up now hold key roles at Apple, Amazon, Alphabet and others

In 2012, a group of entrepreneurs and academics in Cambridge, Mass., hatched a plan to solve one of the holy grails in health tech.

Their idea, which seemed like science fiction at the time, was to create a wrist-worn device to monitor a user's blood pressure and other vital signs around the clock. Uncontrolled high blood pressure affects one-third of Americans, and is linked with heart disease, strokes, kidney disease and all sorts of other complications.

If they succeeded, they could earn billions and revolutionize the new field of wearable computing.

Their start-up, eventually known as Quanttus, attracted some of the most talented engineers, designers and scientists in the country. In less than four years, they designed a watch with an analog set of hands — it was intentionally designed to look like a classic timepiece — with an accelerometer and a heart-rate sensor nestled in the back.

But the data from early clinical trials wasn't good enough, and the company ran out of money before its idea could be proven and taken to market. In spring 2016, it shut down for good, and its last few employees dispersed.

While Quanttus is dead, the dream lives on. A surprisingly large number of key employees from this failed start-up went on to join the health care teams at the world's largest consumer technology companies.

Notable alums include co-founder David He, an MIT researcher who went on to become the technical lead at Alphabet's health tech group, Verily, bringing at least three top engineers with him. His labmate on the academic project, Eric Winokur, did not join Quanttus but went on to join Apple as an engineer behind the original Apple Watch heart rate monitor.

Then there's Yoky Matsouka, a short-lived former Quanttus CEO, who went on to the Apple health team in 2016 before joining Alphabet's home-automation division, Nest, as its Chief Technology Officer the following year. Jordan Rice, who managed hardware at Quanttus and ran the team behind the watch, is now a senior hardware leader at Nike. Maulik Majmudar, the resident cardiologist who helped lead Quanttus clinical studies, went to Amazon in August for a new health-tech bet. Euan Thompson, a board member and interim CEO, went on to become the head of digital health at Samsung and is now at Johnson & Johnson. Other alums joined Facebook, Microsoft, Fitbit, AthenaHealth and the Chan Zuckerberg Initiative. The list goes on and on.

Many of these companies now have blood pressure projects of their own.

Verily, Microsoft, Apple and Samsung all have teams working on the problem, and have been dabbling in the blood pressure space for years, according to multiple sources familiar with their internal projects. (Verily and Apple declined comment. Microsoft did not return a request for comment.) Apple, Microsoft and Samsung have all applied for one or more patents in the space, too.

Samsung has been the most public about its efforts, as it has a blood pressure-related partnership with UC San Francisco. "Sensor technology is advancing quickly," said Jack Ahn, vice president of Samsung health strategy and R&D, in an interview. "We are paying attention and we want to bring that capability (continuous blood pressure monitoring) to the wearable."

The first to crack it will be richly rewarded. A May 2017 report from Grand View Research puts the market for blood pressure monitoring at nearly $12 billion by 2025. Other investors in the space say it's more like $40 billion, if you include all the potential applications for more continuous monitoring with a wearable.

If these tech giants succeed, Quanttus may turn out be the most influential failed tech company since General Magic, which anticipated the mobile revolution in the early 1990s and boasts alums like Android co-creator Andy Rubin, iPod creator and Nest co-founder Tony Fadell and eBay founder Pierre Omidyar.

To learn more about Quanttus and its tantalizing brush with success, CNBC spoke to nine former Quanttus employees and boardmembers, some on the condition of anonymity as they are not authorized by their current employer to talk to press.

They shared new insight into the company's wild history, including high-level talks with Fitbit to buy the company shortly after its IPO in mid-2015. Others spoke about secretive meetings with Apple, Fitbit and Samsung to discuss potential collaborations, investments, and other opportunities, which gave employees a reason to believe that they were onto something big.

Like many medical-tech start-ups, Quanttus formed around a project that started in an academic lab.

Around 2012, David He, a newly-minted PhD from MIT, hatched a theory. He thought it would be possible to measure blood pressure using a light-based optical sensor and a ballistocardiogram, or BCG, which measures the tiny vibrations of the body as the heart pumps blood. Shahid Azim, a serial entrepreneur, and Richard Bijjani, a veteran technologist, rallied behind him.

He had created some noise in the academic community with an early prototype of a device that could measure the heart's electrical signal from the ear. But He's driving passion was to help people with high blood pressure.

One of his inspirations was a set of vintage photographs of low-friction beds from the 1950s, where doctors measured the body's recoil as the heart beats as a proxy for cardiac function. "If you hold out your arm, it's not stationary, it's moving," He told CNBC. 'That's a signature."

The timing seemed right. In 2012, accelerometers that tracked movement were becoming increasingly sensitive. Quanttus' founders, including Azim, had close relationships with chip makers who were making these sensors and expressed interest in medical applications.

As the trio got more serious about the idea, Azim started talking to his venture capitalist contacts out west. Very quickly, he caught the attention of Vinod Khosla, a Silicon Valley-based investor who was increasingly dabbling in health-tech with investments in companies like Jawbone and AliveCor.

Khosla found the vision compelling enough to lead a small seed round of about $3 million, turning the academic idea into a start-up.

"I met Vinod the first week of December in 2012 and basically chased him down at a conference with the elevator pitch," recalled Azim. "We ended up spending more than an hour with him."

With that funding secured, Quanttus started recruiting some of the best hardware designers and product thinkers they could find. Many early employees joined because of the mission and the equity package. In a few cases, Khosla made a phone call, convincing prospects of the eventual impact from the company.

Word spread, and soon the big tech companies started to sniff around.

In fall 2013, three senior leaders at Quanttus flew out to Cupertino to meet with a friend at Apple and share some high-level details about what they were working on with their wrist-worn device. It was intended as an informal catch-up, nothing more. At that time, Quanttus had only about six employees and they hadn't shared anything publicly about their plans.

When they arrived, a staffer ushered into them conference room brimming with about 30 Apple employees, who fired off question after question about their blood pressure technology.

The two companies never formed any official partnerships, but employees on both sides recall regular meetings. One person recalls sharing some early data with Apple, and another described confidential talks about a joint development project related to blood pressure. While Apple remained mum about its plans, as it typically does, Quanttus shared some data and kept the company apprised of its key milestones.

In September 2014, about two years after the meeting, Apple released the Apple Watch. It didn't include blood pressure monitoring, but Apple did emphasize health features like fitness tracking.

The early Quanttus team also paid regular visits to Samsung. One of the people said Samsung offered a letter of intent around an investment in 2015, but the board rejected it.

Fitbit's CEO James Park also came close to buying the company, some of the people said, but both sides couldn't agree on the terms. (Fitbit declined comment.)

While Apple and other tech giants courted the founders, the product team was trying to design a smartwatch that would appeal to fashion-conscious seniors with high blood pressure.

A sporty design wasn't what they had in mind, and anything too techie would be a definite turn-off.

Instead, they came up with a device that looked like a fancy watch, including a dead-fronted set of OLED displays behind the watch hands. The displays turned off until the user interacted with the device, so it looked like a regular analog timepiece.

It was so far ahead of its time that former employees say it was challenging to find suppliers who could execute on that vision. The closest thing to it is a watch that LG released in March 2018, years after the Quanttus design was hatched.

The hardware team faced another big challenge with battery life.

Blood pressure is known to decrease at night, but it's almost impossible to track with a cuff that is programmed to inflate and deflate every twenty minutes — it's bulky and annoying, and wakes the patients up. But it's useful to measure blood pressure fluctuations overnight, as people who experience a less-than-normal nocturnal dip, known as non-dipping, have a higher risk of cardiovascular problems.

That meant that any device would need a battery that lasted at least 24 hours, so it could stay on the user's wrist overnight.

The team at Quanttus tinkered until they came up with a battery that could last for about four to five days of continuous sensing, according to people familiar. They ultimately got it to work well for a small-scale beta release, and had plans to improve that still by incorporating some of the lower power chips coming onto the market.

By way of comparison, today's Fitbit fitness trackers last five to seven days on a charge. The Apple Watch can last almost two days, but it's doing a lot more than tracking vital signs -- it's also sending and receiving communications signals, connecting to the internet, and so on.

While the hardware team wrestled with the design, Quanttus faced much bigger challenges around the science. Tracking blood pressure with an always-on wearable device turned out to be harder than anybody expected.

For starters, the entire question of how to measure blood pressure is controversial. The scientific community disagrees whether the "gold standard" should be measurements from an inflatable cuff, which can be done with inexpensive home cuffs, or an arterial line, which needs to be done at the hospital.

Patients who use an inflatable cuff are required to sit still. But a wearable would need to show accurate measurements as the person moves throughout the day. And blood pressure is constantly changing, depending on whether the user is exercising, stressed out, eating and so on.

Turning the device into a watch was also a challenge, as blood pressure is most accurately measured with a cuff that's close to heart level — around the upper arm, not the wrist.

Making matters even more complicated, the human body varies a lot. Some arteries are more rigid than others. Some people have more fatty tissue around their arm, some have less.

Whatever the reasons, when the company started doing clinical studies in a broader population of a few hundred people, the results simply weren't good enough. International standards from groups like the AAMI accept an error of no more than 5 mm Hg and a standard deviation of no more than 8 mm Hg. (Some investors who follow the space, like Menlo Ventures' Greg Yap, will ask entrepreneurs how close they are to the elusive "5 and 8.")

Sources differed on how close Quanttus got, but all agreed that the results were inconsistent.

As one former employee put it, "When it worked, there was some signal, but when it didn't, we could never really figure out why." In about five to ten percent of cases, they said, the algorithm was way off.

By 2014, the company was engaged in an internal debate about whether they should release a first version of the watch with lower ambitions. The hardware and design teams were rightfully proud of their work and wanted to see it out in the world.

Some thought it might do well as a more basic fitness tracker, tracking easier information like steps, sleep and heart rate, while the team quietly continued to work on the holy grail of blood pressure. The company could collect data from early customers to help in its quest.

But former employees say the start-up's board of directors wanted to stay focused on potential medical uses for the device, and that meant blood pressure was critical.

Quanttus' then-CEO Azim regrets that decision and thinks the company could have benefited from keeping more watches into users' hands. For instance, he says, one former employee found out he had an undiagnosed medical condition called sleep apnea by wearing the Quanttus watch at night.

"We needed real world data from users in the wild," he said. "We had the team to solve it but we needed more time."

In spring 2015, Azim left and a hunt began for a replacement leader. The board chose Yoky Matsouka, who was considered to have the necessary academic reputation and star-power from her background at Nest and Google to usher in a turnaround. Khosla also funded the company with an additional $10 million or so bridge round, which Quanttus did not disclose.

At that point, the company was burning through its cash reserves quickly and losing key talent, mostly to bigger technology companies. The company struggled to raise yet more money, as it didn't have the data it needed to make a compelling argument to investors. Matsouka left the company after just a few months.

The company didn't last long after that. The new chief, Allison O'Hair, set about creating an app for logging health data measurements called Q Health, but that was a far cry from the original vision. The board wasn't convinced that the team could solve the problem of monitoring continuous blood pressure quickly enough, especially after outside consultants reviewed the data. In the end, they decided to shut the company down and write off the investment. Quanttus closed shop in spring 2016.

"It's a very hard problem and a very valuable problem and desperately needed," Khosla told CNBC. "In the case of Quanttus, the technology didn't work and at some point, we made the decision to let it go. It's the nature of hard technical problems that very often, they fail."

Many of the big issues that stumped Quanttus are still unresolved.

Start-ups still struggle to decide: Should they release a fitness tracker first and risk competition from larger companies using identical technology and competing mainly on price and marketing budgets? Or, should they go for the holy grail and try to get validated for medical uses, and risk failing like Quanttus if the money runs out?

Bigger companies like Apple and Fitbit have started with lifestyle trackers while working on medical uses in the background. Apple, for instance, is planning to release a medical-grade EKG sensor in its fourth edition of the Apple Watch in the coming months. People familiar say it's been working on the technology for years.

Verily, meanwhile, seems to be more focused on getting right the medical validation for its wearable, the so-called "study watch." It is currently being used in its clinical trial called "Baseline" but isn't for sale to the general public.

Meanwhile, research continues on how best to measure blood pressure with a wearable.

There's still a lot of interest in pulse transit time and pulse arrival time — the theory that there's a link between a pulse wave arriving at distant measurement point and blood pressure. But off-the-shelf electrical sensors struggle with latent period between when the electrical signals go to the heart to tell it "pump," and when it actually pumps.

Other teams have subsequently looked at using radars and tiny mechanical pressure sensors as a way to solve the problem.

Catherine Liao, the co-founder of a blood pressure monitoring start-ups called Blum.io, has been working on the problem for about four years.

There's a big market, Liao explains, in helping people understand how they're responding to medications. Even people with hypertension get their blood pressure checked only a few times a year. So doctors don't have adequate insight into whether they should change the dosage of blood pressure medicine, or take patients off the medicine entirely and emphasize lifestyle changes.

"We currently know so little because the modality for a more continuous measurement and there are so many things that blood pressure responds to," she said.

Menlo Ventures life sciences investor Yap points out, "Blood pressure is probably the best validated clinical measurement ever, but continuous blood pressure data has never been available at scale," he explained.

"I mean, we thought we could predict heart attacks," added Quanttus investor Stan Reiss. "And who wouldn't buy that device? I'd absolutely be wearing one."

Lead investor Khosla agrees. "It's a very hard problem and a very valuable problem and desperately needed," he said. "If I saw great tech, I'd invest in it today."

Let's block ads! (Why?)

from Top News & Analysis https://ift.tt/2SYZBPf

No comments:

Post a Comment