Edit

Share via

What is Azure Native Qumulo for picture archiving and communication system vendor neutral archive?

  • Article

Azure Native Qumulo (ANQ) Scalable File Service can provide a picture archiving and communication system (PACS) with a vendor neutral archive (VNA) as a solution for healthcare providers. Healthcare providers can then store, use, and archive medical images and other data from one or more vendor-specific PACS platforms.

This article describes the baseline architecture for deploying a Picture Archiving and Communication System Vendor-Neutral Archive (PACS VNA), using file storage services provided by Azure Native Qumulo. The ANQ backed PACS VNA provides a cost effective storage solution with exabyte-plus data scalability and throughput elasticity.

Hybrid architecture

This hybrid solution consists of an on-premises PACS solution, paired with one or more imaging modalities (for example, CT / MRI scanners, ultrasound and X-ray devices) and connected to an on-premises Qumulo cluster. The Azure-based portion of the solution comprises one or more VNA application servers, deployed in the customer’s Azure tenant and connected through VNet injection to the ANQ instance, which is hosted in Qumulo’s Azure tenant.

Note

Qumulo has no access to your data on any ANQ deployment.

Solution architecture

Here's the basic architecture for ANQ PACS VNA:

Access between the customer’s on-premises resources and application and data services on Azure is provided through either an Azure VPN Gateway instance or through ExpressRoute connection.

Application-tier services are deployed to a virtual network in the customer’s own Azure tenant.

The ANQ service used in the solution is deployed in Qumulo’s Azure tenant.

Access to the ANQ cluster is enabled through VNet injection from a dedicated subnet in the customer’s Azure tenant that connects to the customer’s dedicated ANQ service instance in the Qumulo tenant.

Conceptual diagram that shows the solution architecture for pacs using Qumulo.

Solution workflow

Here's the basic workflow for Azure Native Qumulo PACS VNA:

  1. All data in the solution originates on-premises in the form of image files, generated by the customer’s imaging modalities and ingested into the customer’s PACS solution, where they're used for patient treatment and medical case management.
  2. The process of moving image data from the customer’s proprietary on-premises PACS solution to an Azure-based VNA environment begins with the export of image files from the PACS platform to an on-premises Qumulo cluster.
  3. From there, the image files, formatted using the DICOM image standard, are migrated to an Azure Native Qumulo instance using Qumulo’s replication engine.
  4. Access to the VNA imagery is enabled through one or more VNA server virtual machines in the customer’s Azure tenant, connected through either SMB or NFS to the ANQ deployment. Note: ANQ can share the same data in the same namespace through SMB, NFS and object protocols. With ANQ, there's no need to manage duplicate datasets for Windows vs. Linux clients.
  5. VNA imagery can be viewed through any DICOM-compatible Universal VNA Viewer client, located either on-premises or remotely.

Process flow

The process flow for Azure Native Qumulo for PACS VNA is depicted here:

Conceptual diagram that shows the process flow for picture archiving and communication system with a vendor neutral archive using Qumulo.

Components

The solution architecture comprises the following components:

Considerations

Enterprises planning an Azure-based PACS VNA solution using Qumulo should include the following considerations in their planning and design processes.

Potential use cases

This architecture applies to healthcare providers and other healthcare organizations whose PACS data needs to be accessible beyond the originating PACS application.

Your enterprise can use this solution if you're looking to satisfy any or all of the following applicable scenarios.

  • Healthcare providers that use multiple imaging applications and use a standardized imagery format to improve clinical workflows and continuity of care between different systems.

  • Healthcare and life-sciences organizations that conduct research or analysis using medical imagery from multiple sources, whether for short-term or long-term studies that span decades of required archive storage

  • Healthcare and life-science organizations that share documents and imagery with outside providers and organizations.

Scalability and performance

Enterprise architects and other stakeholders should ensure that their solution addresses the following scalability and performance factors:

  • Capacity and growth:

    • a physical Qumulo cluster can be expanded to add capacity as needed. However, for best performance, the on-premises cluster should be sized with sufficient capacity to support 3-6 months of data intake and retention for both the PACS and VNA datasets.

  • Performance and throughput:

    • The on-premises Qumulo clusters should be configured to support the intake of data from the PACS application(s) while simultaneously replicating VNA data to the ANQ target. The ANQ service should be able to receive incoming replication traffic from the solution’s on-premises Qumulo clusters while concurrently supporting all data requests from the customer’s VNA Application Server.

Security

Although each ANQ service instance is hosted in Qumulo’s own Azure tenant, access to the data on the cluster is restricted to the virtual network interfaces that connect through VNet injection to a subnet in the customer’s Azure tenant. All ANQ deployments are inherently HIPAA compliant, and Qumulo has no access to any customer data hosted on an ANQ cluster.

All data at rest on any Qumulo cluster, whether on-premises or ANQ-based, is automatically encrypted using a FIPS 140-2 compliant software algorithm. Qumulo also supports encryption over the wire for all SMB and NFSv4.1 clients.

For all other aspects of the solution, customers are responsible for planning, implementing, and maintaining the security of the solution to satisfy all applicable legal and regulatory requirements for their industry and location.

Cost optimization

Cost optimization refers to minimizing unnecessary expenses while maximizing the value of the actual costs incurred by the solution. For more information, see the Overview of the cost optimization pillar page.

  • Azure Native Qumulo gives you a choice of multiple capacity-to-throughput options to meet your specific workload needs. As the workload increases, the service automatically adds more bandwidth as needed. ANQ adds throughput capability in 1 GB increments up to a maximum of 4 GB. With ANQ storage capacity, you pay only for what you use while you use it.

  • Refer to your preferred VNA server vendor’s solution documentation for specific guidance on virtual machine sizing and performance to meet the required performance for the solution.

Availability

Different organizations can have different availability and recoverability requirements even for the same application. The term availability refers to the solution’s ability to continuously deliver the service at the level of performance for which it was built.

Data and storage availability

The ANQ deployment includes built-in redundancy at the data level to provide data resiliency. To protect the data against accidental deletion, corruption, malware or other cyber attack, ANQ includes the ability to take immutable snapshots at any level within the file system to create point-in-time, read-only copies of your data.

While the solution uses Qumulo’s replication engine between the on-premises cluster and the ANQ target, its purpose is to move data from an on-premises source to an Azure-based target rather than as a means of data protection.

Both ANQ and the on-premises Qumulo cluster also support any file-based backup solution to enable external data protection.

Resource tier availability

For specific information about the availability and recovery options for the VNA application server, consult your VNA server vendor’s documentation.

Deployment

Next steps