Web Exclusive

What Radiology Needs to Know Before Moving to the Cloud

By Steve Deaton

Many industries have shifted to the cloud for increased security and greater efficiency, yet radiology has been slow to adapt. Why has it been so difficult for radiology? There are critical tools that can help the industry expect more from its cloud vendors and IT infrastructures. Other remarkably efficient industries depend on the cloud for secure access to store and manage their data—and those practices can be adapted to radiology.

Radiology is unlike other industries in that it manages extraordinarily large image files, such as cardiology and mammography exams. In addition, radiologists cannot risk making changes to infrastructure that could impede their workflow or efficiency. However, many radiology groups are facing costly fines for security breaches occurring with some types of server solutions. Can radiology really afford legacy systems to support their infrastructures?

Although relatively new to health care IT, multitenant cloud is already being used in other industries that also depend on high efficiency and securing personal information, such as banking. Industries like this are using graphics processing unit (GPU)–powered multitenant cloud, which is different from typical cloud solutions that rely on displaced hardware. Typical systems are difficult to keep updated and are more vulnerable to breaches, given the multiple entry points. Multitenant cloud leverages GPU-powered software that has built-in redundancy and robust security protocols, ensuring uninterrupted and secure access to images.

For radiology groups to feel comfortable making disruptive changes to their infrastructures, they need to be confident that the disruptions are worth it and new processes and infrastructures are secure, efficient, and cost-effective. To thoroughly evaluate a cloud-based PACS, radiology departments should be equipped to challenge their solutions provider with the following requirements.

When thinking about security, reverse the question to consider what makes traditional systems vulnerable to breaches. A leading cause of data breaches is having multiple entry points to secure information. A best practice approach in preventing cyberattacks is to reduce the target size; fewer physical servers (access points) equates to fewer opportunities to expose data. Many facilities today are facing outrageous fines due to patient data leaks that could have been prevented.

Another way to ensure your data are secure is requiring data encryption—when at rest (on a server) and while in motion through a secure sockets layer interface—so that data are never truly exposed. Also, software hosted on the cloud has built-in redundancy and more robust security protocols, ensuring uninterrupted and secure access to images. Many of the best practices used for personal banking, web e-mail, and customer relationship management software using permission-based security can be applied to multitenant cloud PACS for radiology, enabling more data security than is provided on legacy systems.

Another hesitation by radiology groups is concern about efficiency. With an influx of patients over the past year and even in “normal” times, radiology is inundated with not only the volume but also the size of images; there is no room for a lag in efficiencies that they have worked so hard to optimize.

How do cloud technologies support or hinder optimal efficiencies?

First, there is a big difference between a GPU and a central processing unit (CPU), both of which are used to process data. One GPU can handle 1,000 times the amount of image rendering work as a single CPU, which means industries such as radiology need the horsepower of GPUs to handle the enormous images they work with every day. CPUs are designed to handle various generalized data, making them optimal for smaller datasets.

In addition to the horsepower of GPU technology, AI-enabled processing addresses latency issues for large file transmissions, such as the ones used in radiology. AI-enabled processing prioritizes image loading based on the user’s action. For example, the system understands what series a user is reviewing and prioritizes image processing immediately. Tasks such as 3D volume rendering of CT scans can be liberated from dedicated or fixed workstations for diagnostic purposes.

When evaluating a solution for efficiency, ensure your provider can meet or exceed 99.99% uptime and guarantee enterprise-level redundancy, as well as offer GPU technology. These considerations will help ensure minimal disruption when moving to the cloud and thereafter.

The third important factor to consider when moving to the cloud is cost. Although the upfront investment can seem more costly, cloud technology eliminates the need for ongoing one-to-one hardware maintenance and updates. Multitenant cloud (shared hosting) generally provides cost savings, since the costs are shared among all participants, as opposed to individual servers (non-multitenant). Also, cloud-based technology eliminates the need for expensive servers to store data. The ongoing costs of one-to-one maintenance and support for traditional physical servers escalates quickly with one-off updates and upgrades, which also expose more security vulnerabilities.

The cloud is ready for radiology. However, not all cloud solutions are created equal, and it can be difficult navigating the differences. Radiology groups moving to the cloud should seek vendors who offer multitenant cloud to effectively manage and store the large datasets inherent within radiology as well as eliminate possible security vulnerabilities. Finding the right fit for an organization is the key to ensuring that workflow is enhanced, efficiency is optimal, images are secure, and cost is manageable.

— Steve Deaton is the CEO of EvoHealth.