This week is the annual gathering of the RSNA -- the Radiological Society of North America. And for many of us, it's not just another industry trade show: it's a microcosm of the powerful forces that will inevitably play out in industry after industry as we move to pure digital business models.
The underlying premise of a digital business model is simple: it's a complete re-invention of the value proposition along digital principles vs. simply automating familiar physical activities. And, in health care, the business at hand is delivering better patient outcomes far more effectively and efficiently.
Within the broad spectrum of healthcare, radiology in particular is at ground zero for this transformation. Modern radiology has become an entirely digital proposition for the most part, and -- in the process -- radiology itself has been transformed.
And, like so many things, there are lessons to be learned from what's happening in radiology -- not only within healthcare itself, but more broadly when looking at a wide range of professions that haven't been sucked into the digital vortex – yet.
The Other Kind Of Big Data -- Big Content
So much of the big data discussion these days is focused on the amazing potential of predictive analytics powered by new rivers of big data. While much of this attention is justified, there's another side to the discussion -- big data content.
In these models, simply powering analytical feeds isn't the primary goal; the data itself has obvious and recognizable intrinsic value -- the more, the merrier.
I see radiology as a textbook example of a profession now being radically transformed by big data content.
Examples of other professions can easily be spotted; not only within healthcare, but also in industries as media and entertainment, military applications, surveillance and a growing cadre of research pursuits. Some professions -- like energy exploration -- have already been transformed.
In any digital world, content is king. And many seemingly unrelated professions are now starting to realize that they are the new content kings.
As a result, the "art of the possible" is now a moving target -- a continual tug-of-war between new (and unfamiliar) demands -- and the vendor community running very hard to keep up with what's now being expected.
Let's Take A Quick Tour
As you unpack the various dynamics impacting radiology today, there's more than meets the eye.
A good starting point would be the familiar x-ray. In most clinical settings, these are now digitized, and -- since they're static, isolated images -- they tend to consume comparatively modest amounts of storage.
For example, a relatively humble 50 TB array would be sufficient to support a good-sized traditional PACS (picture archiving and communication system) environment -- if x-rays were all that were involved.
Ah, but there's more to it than that -- much, much more.
Not surprisingly, all imaging data is expected to be preserved, protected and available with minimal delay, which often results in a second copy at a remote location.
Performance matters as well -- from storage through application through network to display device.
Fortunately, patient imaging records tend to exhibit good locality of reference, meaning that various storage tiering schemes can work well to deliver both exceptional performance and somewhat acceptable economics.
As a result, radiology professionals are starting to care more about traditional IT infrastructure issues than in the past — as the performance of the plumbing directly impacts their work.
Throw in requirements around privacy, auditability, recoverablity, etc. -- and it's clear that more is desired other than just your garden variety commodity disk drive farm.
More Data? Yes, Please
But modern radiology isn't just about static x-rays anymore. For one thing, newer approaches inevitably involve long sequences of multiple images -- the more, the better.
The current "big four" in terms of data footprint are CR (computed radiography), CT (computed tomography), MR (magnetic resonance imaging) and MG (digital mammography).
An interesting factoid: a single CT study can routinely involved 10,000 or more images. A several-year-old study at the Medical University of South Carolina showed that the "big four" of radiology techniques were already responsible for 98% of PACS storage capacity.
The capacity numbers are starting to look more exponential, and less linear as a result. Frost & Sullivan recently estimated that US-based PACS capacity grew at average of 20% annually over the last five years, and reached 27 petabytes in 2011.
And that's before the next big capacity waves -- already visible on the horizon -- start to wash over the supporting IT functions.
Maybe 3D hasn't been a success in home television, but it's certainly finding its way into radiology.
While cardiology has already been doing 3D for a while, the FDA recently approved 3D digital mammography as a valuable technique to reduce false positives.
In this one example, 3D images create 8.5x more data than their 2D predecessors. And that's before you put them in motion.
Not surprisingly, there's a broad consensus that 3D imaging techniques will find their ways into more aspects of radiology before long, with a corresponding impact on the amount of data that will be generated, managed and consumed.
That alone would make for an interesting story, but there's more.
It's Not Just Radiology Anymore
Historically, the role of the supporting PACS system was to support radiology. But so many other health care disciplines are now starting to generate their own rivers of rich imaging data that inevitably want to find their way into a shared PACS system. Cardiology studies, orthopedics, physical therapy -- even pathology. The amount of data generated by these other disciplines can easily dwarf the comparatively modest needs of the radiology function.
This interesting twist puts modern radiologists in a rather sensitive position: while they remain the primary and historical power users of PACS technology, they now realize they can play a role in supporting other important emerging uses within a broader healthcare setting.
In one sense, they're becoming content brokers for other medical disciplines.
The RSNA show is expected to attract +60k attendees this year. And I would bet that a significant number aren't practicing radiologists.
More "Data Customers" As Well
PACS images are now an integral part of every patient's modern health care record.
That means, simply enough, any entity who is entitled to view those records will inevitably expect convenient and expedited access to all that rich imaged data -- presented in the overall context of a patient's record, and not via some separate byzantine process.
Certainly, that maxim applies to anyone providing healthcare services and not just radiology professionals: whether they practice at the facility where the study was originally done, or perhaps another unaffiliated healthcare provider.
A quick story? Here in the Boston area, we're fortunate enough to have an unusual density of very elite healthcare providers. If you're planning to have a very nasty healthcare issue, Boston is a very good place to do it.
As a result, patients are frequently referred across different facilities as there's a plethora of specialty providers. The patients move easily, but their data doesn't. The stories you hear about the difficulties involved of just getting your medical records from one place to another -- well, they're legendary.
Very often, expensive and difficult radiology studies are done again simply because that's the easier thing to do.
But it's not just the healthcare providers who want access -- the payers do as well. Think insurance companies, companies who self-insure healthcare -- anyone who's footing the bill. And don't forget auditors and regulators -- not to mention the inevitable litigation that's part of the healthcare landscape.
PACS data is also becoming much more attractive from a research perspective. Whether it's simply investigating the efficacy of different treatments, or a longitudinal study, or maybe as just another data type as part of an improved predictive model -- the output of a radiological study is increasingly seen as an input into a quest for better medical treatment.
The picture should be clear: even though the data was captured by (and primarily used by) radiology professionals, everyone is clamoring for access. And those demands will only increase over time. Radiologists and their PACS environments have become the content platform for all.
Don't Forget The Patient!
My local healthcare provider is fairly progressive. For example, they've got a nice web portal where I can review my records and doctor's recommendations, renew prescriptions, ask simple questions, download my healthcare records in a portable format, etc. -- all without calling or scheduling an appointment.
I love it. It works for me because it's incredibly convenient -- and anything that's easier to consume inevitable gets consumed more -- which is a good outcome when we're talking about healthcare.
Not to mention it makes me feel like I'm in control and accountable for my own health -- which is a desirable state of affairs.
I believe it works for my healthcare provider because they can drive better health outcomes at a far lower cost than before. They can have more frequent interactions with me because it's all done electronically. And I don't have to struggle to remember what they told me :)
Recently, I had some MRI work done. Strangely, it wasn't on the portal. I could see the doctor's orders, and the terse writeup from the interpreting technician -- but not the MRI record itself.
Strange, I thought. The MRI procedure was not cheap, why wasn't it accessible? I asked my doctor -- what's up? He really didn't have much to say, other than they hadn't gotten around to it yet.
Did I mention that the theme of the 2012 RSNA event is "Patients First"?
A Banking Analogy?
Consider the modern retail bank.
Almost everything happens in digital format these days, including handling such antiquated formalities as paper checks. The bank is held accountable for managing my transactions responsibly. The bank is also beholden to a rich set of regulators and regulations that prescribe what they may and may not legally do. They're expected to produce transaction records on demand, if asked.
But -- ultimately -- my bank is accountable to me. The information on my accounts and activities belongs to me. On occasion, I choose to share it -- with my tax preparer, for example. I don't expect my bank and my tax preparer to have a formal relationship and detailed information interchange protocols.
If my bank couldn't provide me certain kinds of information that I deemed important (and that other banks could provide) -- like important check images -- I'd start to consider finding another bank.
Now, please try to convince me that my MRI records are any different?
Why All Of This Is So Interesting To Me
Consider the sequence of historical events here.
Radiological imaging becomes an indispensable tool for modern healthcare: more images, richer images, performed more often -- it's all about generating and interpreting digital content.
But that same digital content ends up having more value than just what's needed for a radiologist to perform his or her professional function. Other healthcare professionals need to see them, whether they're in the same facility, or somewhere else.
Making matters more interesting, it's not just the radiologists who want to capture, manage and share image data in a shared PACS system. Other medical specialties are increasingly generating and consuming image data. PACS becomes a shared content platform for all.
Medical researchers realize that there's a treasure trove of valuable data embedded in those images, so they collectively raise their hands and request convenient access -- and with good cause.
Regulators and auditors want access. Insurance companies and other payers want access. Don't forget litigation access. And, of course, me as an informed and empowered healthcare consumer -- I'm starting to demand access as well.
All of the sudden, radiology isn't just about capturing and interpreting images anymore. It's about providing a valuable information base that many others can use and benefit from.
This is big data content in action -- multiple producers and multiple consumers of extremely valuable content that's difficult to manage using traditional techniques. And radiologists are finding themselves being drawn into the role of being content platform builders and content brokers.
And, from where I sit, more than a few of the people attending RSNA this week are fully aware of their evolving role, and are starting to look at much broader requirements than before.
Why All Of This Is So Interesting To EMC
Anytime there are mountains of information doing interesting things, you can be pretty sure EMC is involved, and this week's RSNA show is no exception.
You can find us at booth #3301 Hall A in the South Building.
Here's what we'll be showing — please come by and chat if any of this is interesting to you!
#1 -- Transforming PACS Infrastructure
Anyone involved in PACS is acutely aware of the big data bomb exploding in their backyard, so -- of course -- storage is a big topic.
Many healthcare providers gravitate to products like the VNX; but the larger shops are taking a strong interest in scale-out architectures such as Isilon.
In addition to storage, EMC's data protection, availability and security disciplines also are relevant; there's also strong interest in the newer crop of focused IT service providers who can deliver PACS storage (either primary or secondary) as a service -- many of who are EMC partners.
#2 -- Transforming Vendor Neutral Imaging
VNI is understandably a big deal in the PACS community. Although there are some surprisingly productive standards in this space (DICOM, for example), medical imagery is still substantially entangled with the specific application that produced it -- which, of course, limits its usefulness going forward.
It's a big challenge, and EMC is actively working with multiple industry partners to advance the state of practical VNI through a variety of means, and we'll be sharing our latest work at the show.
#3 -- Transforming Healthcare Mobility
In most any healthcare setting, the professional staff is almost always moving around -- they don't want to be rooted to a desktop to do their work. Enter a particularly demanding set of VDI, mobile app and back-end infrastructure requirements that are distinctly different than what you'll find in most enterprise environments -- the least of which is to deliver medical imagery via VDI :)
At the show, we'll be sharing the healthcare mobility solutions we've built for our customers, using all of our capabilities to deliver the demanding set of expectations in these environments: performance, availability, ease-of-use and security.
#4 -- Transforming Patient Outcomes
Perhaps the most interesting work we're doing is in the analytics space -- using big data content to drive big data analytics models. Medical imagery has shown to have the potential to be powerful fuel for better predictive models, and ultimately deliver better patient outcomes.
We'll be demonstrating examples of how we're using our Greenplum capabilities to make researchers even more productive in harnessing vast amounts of rich content to uncover new patterns and insights.
A Bigger Picture?
While it's true that every industry and every profession is understandably unique, it's not hard to spot patterns that emerge in one sector, and quickly propagate to seemingly unrelated activities.
Here in the US, it's not hard to make a case that healthcare is being completely transformed through digital technologies. It's not about simply making familiar processes more efficient; it's a thorough re-thinking of how patient outcomes are improved while at the same time dramatically improving efficiencies and reducing costs.
In many aspects, radiology is where it's happening first within healthcare. Their rapid move to purely digital models has not only transformed their profession, but has also broadly expanded their role within the healthcare community.
As with all digital models, content is king.
And, within healthcare, radiologists are quickly becoming the new content kings.