I’m most excited about the 3rd and final part of SwiftStack’s blog series on IT strategies for Medical Imaging, as we’ll be summarizing a conversation with someone who manages imaging systems for a living. I recently spoke with Dharmesh Patel of Atlantic Medical Imaging (AMI), a radiology clinic with a dozen offices in New Jersey that services area Hospitals. (Disclosure: AMI does not currently use SwiftStack technology, which allowed us to focus the discussion on their environment, challenges, and initiatives.)
In many ways, AMI is a traditional health care IT shop, with a foundation of VMware, Cisco networking and compute (UCS), and NetApp storage. Their PACS system is the core of the operation, and that’s where things get interesting. AMI is transitioning from major PACS vendor A to major PACS vendor B, which Dharmesh describes as a “big pain” for IT. We’ve heard this before about PACS migrations, where the devil is clearly in the details. In AMI’s case, the fact the two PACS systems use different data compression ratios dealt an unexpected blow to the project budget: they need more storage to complete the migration.
There’s Just Not a Perfect Product
The logical answer to PACS incompatibilities would seem to be a Vendor Neutral Archive (VNA), but here again, there can be gotcha’s. To overcome these limitations, many VNA’s require that patient images be captured directly from the modality to provide universal viewing to physicians, one of the major reasons for going with a VNA in the first place. So, legacy PACS data must be migrated (there’s that word again) to the VNA to realize the full benefits of an enterprise-wide clinical repository.
Integrating the imaging systems with Electronic Health Records (EHR) software is another area where the potential efficiencies are compelling. When done properly, all information about a patient is available through a single source, with the IT complexities hidden from the clinical staff. In AMI’s case, this EHR to imaging integration is elusive as there are so many different products in the mix, both in their practice and in use inside their Hospital customers. Broad support of middleware standards would certainly help.
Using Dharmesh’s words, “there’s just not a perfect product.”
And the Doctors Say…
Since the physicians own the radiology practice, it’s vital to know what they want from IT with respect to patient imaging. Cost management is at the top of the list, which is not surprising given how fast data is growing in the imaging world. For example, some HD mammography studies are now 5GB each, and data sizes will continue to increase as image clarity improves. Simplicity of access is also a key request. Doctors want to spend less time navigating systems and clicking buttons and more time treating patients. And, the ability to have anytime/anywhere access is an obvious need in today’s 24/7, mobile world where “office hours” is a concept of yesteryear.
AI Gets Real
Utilizing health data for better patient outcomes and the overall good of society is a hot topic, as evidenced by a cover story in the March 2018 issue of Fortune Magazine: Big Data Meets Biology. In the radiology world, artificial intelligence (AI) is gaining traction due to its ability to deliver faster and more precise radiology examinations, with algorithms only improving over time. Much of this innovation is happening in the Public Cloud, but AI integrations with imaging systems – both PACS and VNA’s – is a logical evolution and promising development.
Takeaways for IT Infrastructure Providers
While a 30-minute phone conversation doesn’t present a complete picture of the imaging landscape, there are some common themes relating to infrastructure:
- With the never-ending growth of health data, cost and scale will always be top factors for infrastructure decisions
- Applications drive clinical operations, so infrastructure components must offer flexible interfaces and support industry standards to avoid being yet another silo
- Since Medical Imaging comprises more than just PACS, systems that store/manage images must be optimized for non-DICOM data types
- The rapid technology transformation in health care is bringing AI and Cloud to the forefront, but these advancements must be implemented in combination with legacy systems to deliver real value
At SwiftStack, we’re working with our platform partner, Cisco, to build Medical Imaging solutions that address these critical requirements.
Many thanks to Dharmesh Patel of AMI for giving us an inside look at a fast-growing Medical Imaging shop. Our interview underscores the point that, while workflows and applications vary among industries, all organizations are looking to IT as a primary driver of customer satisfaction, operational efficiency, and competitive advantage.
More about the author:
Greg Govatos is the VP of Strategic Partnerships for SwiftStack. This series is a result of Greg’s work with systems integrators whose primary expertise is developing and operationalizing the technology plans of health care providers.