HealthMyne Put the Band Back Together to Unite Medical Images & Data
Some startups are born out of an epiphany or an accident by a brilliant (or lucky) entrepreneur. Other companies’ origin stories are not as sexy, and instead center on a group of smart people putting their heads together to calculate where market winds are blowing and how to get there first.
Madison, WI-based HealthMyne falls into the latter category. The stealthy healthtech startup was sparked by a conversation between local serial entrepreneurs Rock Mackie, Praveen Sinha, and Mark Gehring while celebrating Gehring’s birthday at a local watering hole two years ago.
When the trio sat down at the Great Dane Pub and started kicking around ideas for their next business, they could’ve chosen almost anything, and observers probably wouldn’t have doubted them. After all, their previous startups have achieved more than $3 billion in combined sales, and two of the businesses netted a combined $250 million in initial public offerings (TomoTherapy, since acquired by Accuray for $277 million, and Emageon, which had previously merged with Madison-based UltraVisual Medical Systems).
The entrepreneurs decided that they wanted to make a play in the healthtech space, given their experience in medical devices and healthcare software, the billions of dollars being poured into the sector by venture capitalists around the country, and the Madison area’s growing strength in healthtech—anchored by electronic health records giant Epic Systems.
“We have a monster in town, and they generate a lot of interest,” Mackie says, referring to Epic. “This is a great climate for healthcare IT.”
Many healthtech companies make software that organizes text-based electronic health records so that hospital and insurance administrators can do their jobs better. But not all healthcare software companies target doctors as their end user, nor do all of them work with medical images—partly because images are bigger, more complicated files, and the companies would need U.S. Food & Drug Administration clearance if their products inform the diagnosis and treatment of patients.
That doesn’t faze HealthMyne. The company is developing a search engine and data analytics product that combines electronic health records and medical images, like X-rays and CT scans. The idea is to allow a doctor to rapidly search a healthcare system’s medical imaging archives, bring up the desired image on any device, and overlay a host of other relevant patient data and analytics reports on top of the image. The ultimate goal is to give doctors quicker access to the information they need to make better, faster decisions, ultimately resulting in healthier patients and more cost-effective care.
The company, which officially formed in January 2013, says it has raised $1.25 million so far, a mix of equity investments from the co-founders and angel investors, a federal research grant, and a state technology development loan. It scored a pilot partnership with the University of Wisconsin Hospital and Clinics to conduct early tests of its software, which is handling more than 500,000 medical images and 85,000 pieces of patient data there per day—some 8,200 gigabytes of daily data—the company says.
HealthMyne has another beta-testing partnership with Moffitt Cancer Center, located in Tampa, FL, Sinha says. The executives are reaching out to other major cancer centers, as well as talking with investors to raise more capital. They currently have a staff of 10 people, including seven full-time employees.
The company will soon apply for FDA clearance so it can begin selling its software, and it aims to start generating revenue in 2016, Sinha says.
But before it could even get this far, it had to assemble a top-notch team. And here’s where the story sounds a little bit like the refrain from the movie “The Blues Brothers,” only in this case it’s a few major players in the Madison startup scene who are putting the band back together.
Gehring, a serial entrepreneur who studied biomedical engineering at Marquette University, and Mackie, a UW biomedical engineering professor emeritus, have known each other since the 1980s. In 1992, they co-founded Geometrics, which developed the Pinnacle 3D radiation treatment planning system. In 1996, they sold the company for $3.9 million to ADAC Laboratories, and in 1998, Sinha—who had met Gehring and Mackie while working as an NIH fellow in the UW hospital’s radiation oncology department—took a product manager job at ADAC. Sinha worked alongside Gehring at the company (which is now part of Philips). Mackie had already moved on to found TomoTherapy at that point.
Sinha, HealthMyne’s CEO, and Mackie, the company’s chairman, wanted Gehring as a full-time member of the team from day one, but at the time he was still heavily involved with a different healthtech startup that he co-founded, Propeller Health. Under CEO David Van Sickle, Propeller and its data-collecting sensors for respiratory disease patients have since gained traction with early customers and investors nationwide. Gehring, who says he prefers working with early-stage companies, gradually reduced his time spent working with Propeller as the company began maturing. He served as a HealthMyne advisor until last fall, when he joined the company.
“Getting back together with these guys is so much fun,” Gehring, HealthMyne’s chief strategy officer, says.
Meanwhile, Sinha and Mackie brought on HealthMyne co-founders Roger Chylla and Hao Wang as chief technology officer and chief privacy officer, respectively. Chylla worked with Sinha and Gehring at ADAC, and the trio later co-founded UltraVisual. Wang previously was CTO of NovaShield, an IT security company Sinha co-founded in 2006.
With this group (pictured above) in place, HealthMyne boasts one of the most experienced, successful, and cohesive executive teams among any Wisconsin startup right now.
“We can complete each other’s sentences,” Mackie jokes.
A track record is no guarantee of future success, of course, and HealthMyne’s execs will need to leverage the credibility they say their previous companies earned them with hospitals and clinics around the country. They point to the UW and Moffitt partnerships as early indicators that the strategy is working. But if they don’t deliver a good product that sustains their reputation and builds healthcare providers’ confidence in their new startup, HealthMyne could be toast.
Well-established electronic health records software providers like Epic, Cerner (NASDAQ: CERN), and McKesson (NYSE: MCK) have built big brands and trust that make hospital administrators and doctors “more inclined to use those services,” says Vishnu Lekraj, a senior financial analyst who covers healthcare services companies for Chicago-based Morningstar.
“That’s going to be a big hurdle for a lot of these smaller firms, these startups, to overcome in order to get entrenched or to build their service out,” Lekraj says.
In other words, HealthMyne needs to prove it has got an idea that will interest doctors and healthcare system administrators nationwide enough that they’ll pay for it—and perhaps pass over big-brand options to do so.
Currently, patient health records and medical images often reside in different software systems, which can make for a lot of extra work for physicians. A doctor studying a scan might think it looks like a patient has emphysema, for example, but if the attached notes from the radiologist don’t indicate that, she’ll have to check health records that might only be accessible at a different computer workstation.
“They’re literally getting up, walking somewhere else, getting a pad of paper, making some notes, and coming back,” Mackie says.
Or perhaps a doctor wants to compare a patient’s condition with a similar case from the previous week. There’s not an efficient way to do that now, Mackie says.
“I call it dark data,” he says. “If you can’t get to the data quickly, it’s like it doesn’t exist. Today, they’re wasting minutes, even hours in chasing down things that should be presented to them more easily.”
Plenty of companies offer software that aims to make it easier for clinicians to access medical images and associated data, including McKesson, Merge Healthcare (NASDAQ: MRGE), Agfa HealthCare, ScImage, and EMC (NYSE: EMC). And more competitors are likely to pop up in the next few years, reacting to federal reforms that are pushing healthcare providers to make imaging results electronically accessible and more closely linked with other patient data.
The electronic health records software sector overall is “getting more crowded,” Lekraj says. But there’s still room for smaller companies to grab a piece of the market, given the healthcare industry’s increased investment in such software and the fragmentation that still exists, with healthcare providers using lots of different electronic health records systems—even among hospitals and clinics that are part of the same health system. “It’s a market that is ripe with opportunity,” Lekraj says.
HealthMyne’s software will have several features that could help distinguish it in the market, including faster access to patient data, the ability to compare data across groups of patients, and “high-level image analytics,” Sinha says. “That’s where we think we can differentiate very well.”
With HealthMyne, not only will the doctor be able to quickly call up all of a patient’s relevant data, but the system will have the ability to run custom queries to compare large groups of patients that share certain characteristics, which can be useful in assessing the accuracy of a diagnosis or the effectiveness of a patient’s treatment, Gehring says.
For example, let’s say a doctor is viewing brain scans of a cancer patient taken six months after diagnosis. The software could pull data on similar patients from that same period in their treatment and present the findings in a chart or graphic laid right on top of the current patient’s brain scan. Data points of interest could include the size of similar patients’ tumors after six months and the survival rate of similar patients treated at the hospital over the past two years.
Current hospital imaging systems “are not really set up to compare across patient boundaries,” Chylla says.
Another feature of the software is a comprehensive timeline of the patient’s treatment, which could show when the person underwent scans, when he received chemotherapy treatment, and when he had surgery. Such a tool could help avoid unnecessary MRI or CT scans, which saves money and, in the case of CT, reduces the patient’s exposure to radiation.
The software will also draw colored lines on the medical scan that trace the contours of different segments of tumors, so that doctors can more easily see if the tumor is growing or shrinking. “We would detect the 3D structure of that tumor and the surrounding tissue,” Gehring says. Then, the software can create a chart with all the tumor’s statistics, such as its diameter, its volume, and how much of it is composed of necrotic, or dead, tissue.
Some existing products can handle the tumor-tracing tasks, but they don’t connect with all the other data and provide analytics like HealthMyne does, the company says.
“The efficiencies for the provider are going to be enormous,” says UW radiologist Patrick Turski, a HealthMyne investor and consultant. “The ability to rapidly integrate both imaging and the clinical medical record findings will be very easy, very automated.”
The biggest barrier to HealthMyne grabbing a foothold in the market will be the cost of its system, Turski says, especially considering the healthcare industry has a reputation for moving slowly to implement new technology.
“I think it’ll be hard for hospitals to want to invest in something which is still new and in development, no matter how innovative it looks,” Turski says. “We’re going to have to be able to show some success in some very large places that have very large volumes” of patients.
The exact details of HealthMyne’s revenue model are still being determined, but it plans to charge a fee per use of the software, which should lower the up-front cost for customers, Sinha says.
HealthMyne could be aided by new hospital reimbursement rules in the Affordable Care Act that pay healthcare providers based on quality measures and patient outcomes, rather than the traditional fee-for-service model, Turski says. That might turn a hospital executive’s return on investment calculus in HealthMyne’s favor, if the administrator believes the software will result in better, less costly care.
But HealthMyne could be hampered, at least early on, by the hospitals’ software systems, which might not have all of the necessary patient data housed on servers in a compatible format that will allow the company’s software to quickly find information and present it in a visual way. HealthMyne’s software won’t incorporate natural language processing, for example, meaning it won’t be able to read the free-form notes that radiologists often attach to patients’ scans.
“Those are barriers that’ll have to be overcome,” Turski says. “But they’re recognizable barriers, and there are recognizable solutions.”
Hospitals are starting to invest time and money into creating “data warehouses” that store all of their patient data, and they’re also “scrubbing” electronic health records to fix errors and turn qualitative reports into quantitative data that can be understood by HealthMyne’s software. “They’ve got these nurses reading these reports and saying, ‘Was the tumor getting smaller or bigger?’—whatever they can glean from that report—and getting it into a structured database where you can search and actually do analytics,” Gehring says.
HealthMyne will be strategic when going after customers, initially pursuing large cancer centers that either have or are investing in the data infrastructure that the startup’s software will require.
“Part of the risk we are taking, obviously, is saying the market is going that way,” Sinha says. “In some places, the infrastructure is there, and the infrastructure will be there more and more as we grow.”
HealthMyne is first targeting radiologists and oncologists, but eventually plans to expand its software to cardiology, neurology, and “the entire field of medicine,” Mackie says.
“What we see is there’s a data deluge coming, if it’s not already there,” Sinha says. “It’s about helping doctors interpret that data better. They don’t have the time. Their job is to take care of the patient.”