New Startup Raises $100M for Portable Ultrasound | Hacker News Reader

New Startup Raises $100M for Portable Ultrasound

technologyreview.com

106 pointschriskanan38 comments

Rothberg is something of a character.

His two DNA sequencing companies (454 and Ion torrent) have both had what I would consider successful exits (solid to Roche and Life technologies respectively). But more importantly both products made their way into the hands of consumers (which is actually rare for DNA sequencing startups). However after acquisition both products failed to gain significant traction (Roche has just shelved 454).

He also engages in a certain amount of showmanship [1], which has as gained him some ridicule. I think he has to be respected to successfully executing in a market where most startups fizzle out though.

I hope this pans out though, it sounds awesome. I really hope this (and other research/medical device) technology makes its way into the hands of normal consumers (not just research labs/medical institutes).

[1] Oh and he has a scale model of Stonehenge in his back yard. http://clonehenge.com/2009/01/17/circle-of-life-connecticuts...

I've worked with Rothberg and he is certainly a different than most people you meet in a day. Incredibly gifted with technological vision, business savvy and marketing skills. He also has the ability to put together an incredibly talented team and keep them focused.

Ion Torrent didn't put Illumina out of business or anything but it did define a new market segment in high throughput DNA sequencing for small to mid size labs. Ion Torrent's PGM is like the PC of DNA sequencing to Illumina's mainframe HiSeq and forced Illumina to respond with their MiSeq machine. I for one am very excited to see the democratization of DNA sequencing as it gets more available to smaller labs and individuals.

Can't wait to see what he and the team pull off on this one as we certainly need a few more breakthroughs in the clinic both in price and ease of use!

While 454 may not have much traction anymore, Ion Torrent is coming out with a new series of devices (the Proton iirc?) and they're competing pretty well with Illumina in the cheap sequencing market. It's something, anyways.

I was thinking of trying to make a smartphone-powered ultrasound awhile back and looked into it, and these devices don't look too complicated. You put a piezo in a little box, and cycle it to send/receive vibration in the form of sound. The software to decode the signal into an image is the hard part. But if it's Rothberg behind this...well, I'll take his word for it being a hard thing to make low-cost.

My only experience with Roche is that they buy promising new technologies and shutter them. First they purchase the product, then they invest, then they purchase the company then they shut it down.

This is really interesting. They certainly have a reputation for killing 454. What other companies have they done this with?

This what it felt like with Combimatrix http://www.evaluategroup.com/Universal/View.aspx?type=Story&... who are still around but had a pivot and change of leadership.

I wouldn't say Roche "killed" 454. Illumina is absolutely dominating the sequencing market. I would say killing something that isn't profitable is a good business decision.

You might not, I might not. But they do have somewhat of a reputation for not executing well on the 454 technology. It has been said that they failed to scale out what was a fundamentally sound platform. That they didn't really push on density to the increase device throughput as much as they could have.

The truth maybe that the technology just has fundamental limitations. But I'd would say that's not objectively clear.

The Illumina (Solexa) platform also had significant limitations, it took a lot of platform development to get it where it is today. In contrast Illumina have executed quite well, providing systematic improvements to the Solexa platform.

A sale is not always a successful (i.e. profitable) exit.

No, but I think in Rothbergs case they were ok. Not by any means stunning.

Ion Torrent raised around 60MUSD according to angel.co. They sold for 375MUSD, with milestone payments making it up to 725MUSD. I've no idea if they hit their milestones or not.

454 was a lot less... I think a few hundred million. But they also raised less IIRC.

Some researchers at wash u engineering built a cheap, phone based ultrasound in 2009:

http://news.wustl.edu/news/Pages/13928.aspx

I did various forms of cardiac ultrasound research for 10 years at the med school. So, I know a little bit about it. It's easy to find your heart or a baby, but knowing enough to make any sort of diagnosis takes a lot of practice. No matter how smart you think you are, you won't be able to order one of these on Amazon Prime and find a leaky valve or a problem with your gallbladder. The images are noisy. Positioning the probe to get a clear view of the structure you want to see is an art. If you're obese, forget about it.

I'm all for disrupting the imaging industry with cheap hardware, but this article and a lot of the comments here show the general lack of knowledge about medicine. Have a look at the comments in the heart attack thread. Some people think they can just buy an ECG and self-diagnose a heart attack. If anyone spent twenty minutes looking at ECG's (or ultrasounds) with a knowledgable tech they'd cringe at how stupid they sounded. Ultrasound is the exact same thing.

Finally, ultrasounds are regulated medical devices. As far as I know, it's illegal to own one.

On a happier note, 3D ultrasound is kind of awesome. If you get a probe that can rotate 180 degrees, it is really easy to make a 3D model of the structures. I did it for a work-study job in 1998 on an old SGI indigo. Sadly, you can get much better pictures in MRI or CT so 3D ultrasound is not really used in the clinic (it's gaining some traction in cardiology, though)

I guess what's new here is the capacitive micro-machined ultrasound transducers (CMUT), ultrasound emitters directly on a semi-conductor wafer. Though I would like to see what we can do with this technology without limitations of mobile/portable.

In terms of portable ultrasound, there are a few existing solutions on the market:

This looks pretty portable (connected to an iPhone too): http://www.mobisante.com/

GE's portable ultrasound seems pretty portable too: http://www3.gehealthcare.com/en/Products/Categories/Ultrasou...

Best of luck to Rothberg et al., I believe there's a huge consumer market for a portable ultrasound done right! The pregnancy market alone is vast, e.g. https://www.bellabeat.com/

The difference with GE Vscan seems to be the 3D reconstruction capabilities. However I am not sure if the 3D is really needed. I always understood the 3D ultrasound reconstruction as a gimmick (at least in obstetrics) to show parent the head of their future baby.

Most of the ultrasound-using doctors that I've talked to in med school seem to agree with you; if there's a real need for 3D imaging today, doctors (in 1st world countries at least) send the patient for a CT. The cost/benefit for 3D ultrasound doesn't compare favorably to CT because it is a much noisier imaging process and it takes just about as long as a CT.

That's why if you Google image search for "3D ultrasound", you get mostly pictures of fetus faces and very few medical conditions.

So, I see this product mostly benefitting the 3rd world, unless they vastly improve image quality and lowered the technical hurdles compared to existing ultrasound equipment. (Which would be awesome, but is a hard problem.)

Disclaimer: have used the Vscan as a student, but am no means an expert in ultrasound.

good luck trying to ct scan a pregnant woman!

True, but you've just emphasized my (and the GP's) point: 3D ultrasound is probably most relevant for pregnant women and 3rd world situations.

Incidentally, besides the abdomen and pelvis, you can safely CT just about any other part of a pregnant woman.

On the darker side, a dangerous social impact that this device would have is an increase female foeticide

http://en.wikipedia.org/wiki/Sex-selective_abortion

Though there is a lot being done against this social evil, it is still very prevalent.

I'm quite relieved to see someone is thinking of this use case and bringing it up in the discussion - a classic example of technology having both a positive angle and a negative one. I can't think of a way to solve this dilemma though .. seems like a slippery slope for us to be approaching.

A few questions.

Why would a cheaper technology lead to an increase in abortions? Don't they just kill the fetus anyway upon realization of the sex if it's undesirable?

Wouldn't it ultimately be safer for the mother, to perform a very early term abortion if that's what she wants, rather than waiting?

If a mother wants to abort at 60 days pregnant (assuming the new tech enables earlier sex identification), why shouldn't she be allowed to for any reason she deems worthy?

Because in countries where this is a problem it's common to have extreme restrictions on use of ultrasound machines. For example requiring two doctors to be present to use it and banning gender identification uses.

If this device is cheap/small enough that illegal clinics could have it, it would make gender identification much more widely available then it is now.

Selective abortion can be dangerous population-wise because even a small gender skew can be hugely problematic.

On a micro level I sympathize with freedom of choice.

On a macro level, population skew is certainly a problem.

Tough call.

Not to start a flamewar, but it really isn't a tough call at all. It's just that you got trapped in this debate-framing message that abortions are about freedom.

You know another problem with legalized abortions? Women that sell them in exchange of suing for child support. [0]

[0]: https://blogs.law.harvard.edu/philg/2014/10/30/who-knows-abo...

Because she may not be the one actually making that "choice", and even if she is directly, it is made under massive cultural duress?

I have an ultrasound that we bought on eBay when my wife was pregnant.

The reason more individuals don't own them is that it's illegal! No company that makes ultrasound machines will sell them to a non-doctor, because of some kind of regulations. I believe they don't want people to think they can diagnose themselves when it actually takes a doctor. Whatever the reason, that's why more people don't own sub-$10k ultrasound machines, and omitting that fact to make the story one purely about the march of technological progress is poor journalism.

> I believe they don't want people to think they can diagnose themselves when it actually takes a doctor.

That is not the only reason. Long-term effects of ultrasound are unknown. Short-term effects are increases in temperature of the tissue being scanned. Improper use can lead to harm.

Not saying they should be illegal to own or purchase by consumers, but you really shouldn't be scanning things for shits and giggles. Remember, not that long ago, x-rays were considered routine and harmless.

Another danger is that, if these became popular in places like India and China, it could lead to the abortion of female fetuses in numbers greater than what is already taking place. It's illegal for pregnant women to get ultrasounds in these countries already.

Hope they destroy the industry.

10 minutes of ultrasound to diagnose a serious illness can cost several hundred dollars in the USA

Maybe in a decade it can even be built into smartphones like a tricorder.

This device already looks like it can plug into a smartphone http://www.mobisante.com/products/product-overview/

An you can feed a software consultant on 1 dollar a day.

The price in medicine is mostly not the hardware but the required expertise to analyse the data. Ultrasound imaging is not easy. The american health system is very expensive though, a ultrasound is around 50 euro's here.

That said, 3D ultrasound can have tons of other applications not just in diagnostic medicine. Real time surgcial feedback in soft tissue, yes please!

This is the sort of tool that will find loads of uses in different areas if it's cheap, portable and flexible.

I can imagine engineers using it to examine materials, plant scientists looking at plant growth or builders using it to find out where different things are located in walls.

> 10 minutes of ultrasound to diagnose a serious illness can cost several hundred dollars in the USA

Chances are only a small portion of that is the cost of the ultrasound machine.

Yes you would think they expect to make their $100M back, and you don't do that 'destroying an industry'.

Also their raise ($100M) is an astronomical seed round, 100x what most companies would raise pre-product and pre-revenue in a first round. They clearly expect to deliver at volume on existing very high price points, or would not have raised this much money.

This isn't a few University kids disrupting an industry raising a few hundred k, this is an industry regular opening a division in it with $100M of his own money and that of a few investors.

Exciting development from R&D perspective for sure. Love the infusion of capital and the media buzz.

We have been tackling the opportunity of point-of-care ultrasound with a phone and tablet solution for a few years now (www.mobisante.com). There are interesting nuances related to business model, education, and market adoption even within US, and non-trivial complexity in purchasing dynamics and politics when it comes to emerging markets. These are just as important to break through as cost and complexity barriers – which we have already done to some extent.

That said, we would love to source these transducers when commercially available and cleared by the FDA, and integrate them in our solution offering!

Sailesh Chutani, CEO Mobisante

GE has a $9k version (Vscan), and Rothberg says he wants to make this for $6k ("1000 times cheaper [than a $6M MRI machine]" is $6k). Is that really a significant improvement? I guess any price competition is good, but I was hoping for a <$1k version. Everyone should have one of these next to their toothbrush.

Here's some reports from someone who has been using GE's portable ultrasound machine:

http://whyisamericanhealthcaresoexpensive.blogspot.com/2014/...

http://whyisamericanhealthcaresoexpensive.blogspot.com/2013/...

"It was an unexpected and welcome bonus that my patients and their families loved it. I would share the moving ultrasound pictures with them, often having them hold the machine so I could point out how beautiful their internal organs were and what we could see that helped give us a clue about their disease process. Many of these same patients also got full, detailed ultrasounds or other imaging by radiology technicians, but since the technicians aren't supposed to discuss findings with the patients and often they couldn't see the screen, it wasn't nearly as gratifying."

"When I see a patient now, instead of taking their pulse and placing my stethoscope on their chest and back, hearing the vague taps and clunks and bubbles and whooshes of the internal organs I have come to trust are in there, I open the ultrasound machine that lives in my white coat, squeeze a little gel from a tube I keep warm in my pocket...."

"The most common comment I get from patients is, "Wow, that's really cool!" "I agree!" I answer. Then they will ask, "Why doesn't everyone have one of these things?" That is kind of a difficult question. "They're pretty expensive," I usually say. They are. At least for now. The little machines (Vscan, by GE) retail for over $8000, though you can buy them cheaper used or overseas. Physicians balk at spending this amount of money on a piece of equipment. Most of the expensive gadgets we use are owned by hospitals or by our group practices. Musicians, however, who make a fraction of what we do, buy their own musical instruments which often cost in excess of $10,000. I'm not sure the cost ought to be a serious consideration."

I think in parts this can get down to way under $1k. Right now the bottleneck is the ultrasound transducer probe tips/fronts/arrays, which are apparently quite laborious to make. The image processing techniques are pretty standard and appear as open source software in at least one of TI's development kits.

edit: And as another commenter points out, yeah it's illegal to sell these machines or something (blame the Federal Death Administration).

The price should be in the hundreds. According to re/code, "Butterfly’s goal is to launch the device in as soon as 18 months — pending development and regulatory approval — with a price tag of hundreds of dollars, Rothberg said"

https://recode.net/2014/11/02/butterfly-network-raises-100-m...

Would your average person be able to use and interpret the results?

Butterfly Network, one of the companies mentioned in the article, has a panel discussion next week in NYC on artificial intelligence. Panel includes David Ferrucci (one of the creators of IBM’s Watson) and Max Tegmark (who hypothesizes that consciousness is a state of matter that can be mathematically described). More info here: https://4combinator-speaker-series.eventbrite.com

I think a network of sensors is an interesting, like what you see in Extant. A lot of the ultrasound technicians job is to move the wand around to get images and adjust the frequency to see different things. They do also push down a lot. When we had our babies, I never understood why they just didn't drape a bunch of sensors over and run a single scan that calculates everything.

Best of luck to Rothberg! Amazing!

Reply on news.ycombinator.com