Guest

Cisco Wide Area Application Services (WAAS) Software

Cisco WAAS Ver 4.0 Optimizations for PACS and Digital Image Storage

Overview

Healthcare organizations (HCOs) experience the constant pressure of growing storage and patient requirements. Cisco® Wide Area Application Services (WAAS) Version 4.0 provides application acceleration and WAN optimization capabilities to allow healthcare providers to both better centralize costly infrastructure and improve the delivery of imaging and other medical data to distributed locations. This solution overview and performance brief examines the systems that HCOs use and how Cisco WAAS from Cisco Systems® can help improve the performance of these systems in a geographically disperse environment.

Description

HCOs use a picture archiving communication system (PACS) as the primary application to manage digital images. PACS is a combination of application, database management system (DBMS), Web servers, and storage media, as shown in Figure 1. Digital imaging services are expanding exponentially in the volume of patient encounters, the number of specialties that use the technology, and the number and size of images in a study, as shown in Table 1. The expansion of imaging services requires PACS and the supporting infrastructure to scale beyond radiology and across a much larger, more diverse and distributed environment.

Figure 1. Imaging Workflow Has Crucial Stages and Technology Components

Many remote clinics and sites have limited WAN connections for cost or availability reasons. These same clinics may need to transmit images acquired at their sites to the central archive for radiologist interpretation or other purposes. The WAN link is sized for normal transport demands and not designed to support delivery of large image files. Although this approach is cost-effective for most remote operations, it restricts communications and image delivery assurance. Image transport times across small WAN circuits can exceed Digital Imaging and Communications in Medicine (DICOM) latency requirements, thus causing timeouts or even failures. Even though image modality, PACS, and the network are all operating normally, the image is not delivered properly to PACS because of low-bandwidth, high-latency network conditions. The image transmission is poor, and in many cases the image is retransmitted many times because of the performance barrier caused by the WAN-causing delays in productivity and affecting critical patient care activities.

Table 1. Image Availability and Bandwidth Demands

Type of Study

Approximate Size

Type of Study

Approximate Size

X-ray computed radiography

5 MB

Computed tomography

150-1000 MB

Angioplasty

500-1000 MB

Ultrasound

100-500 MB

Nuclear medicine

200-500 MB

Cardiology

5-10 GB

Positron Emission Tomography (PET) scans

200-500 MB

   

Challenge

PACS architectures are designed to serve primarily radiology images within a campus or regional environment. Many healthcare providers have deployed centralized architectures with a primary archive, thus requiring multiple image transfers across the network to support image acquisition, regardless of proximity to the centralized storage and server repository. These architectures commonly do not include any form of caching or optimizations to reduce repeated retrievals of images or similar data. This approach requires tedious retransmissions between application nodes and users-further inundating the network, reducing available bandwidth, and degrading performance.
Select PACS architectures improve image retrieval through distributed short-term storage archives, expanding the image data file system and storage environment. Depending on the vendor, image type, and archiving scheme, the PACS writes copies to both the short- and long-term archives, resulting in multiple image copies traversing the network or protocol overhead for managing the image copies, and also requiring management of multiple silos of infrastructure.
The PACS management protocol, DICOM, is a latency-sensitive transport protocol that uses common file protocols for transfer, including the Network File System (NFS) for UNIX environments and Common Internet File System (CIFS) for Windows environments. Most PACS software works with any file server or NAS device. The bandwidth restrictions of the WAN coupled with the chatty nature of the DICOM transport protocols result in significantly increased transport times, thereby causing transfers to exceed DICOM criteria. This scenario can cause image delivery failure, leading to frustration and reduced productivity.
HCOs have several options to meet these demands and address these concerns:

• Incur higher WAN services fees to support image transport-A solution that does not address the problem of latency or redundant transfer

• Reduce image volumes or limit imaging services-Resulting in lost revenue and productivity

• Limit image transport distances or geographies-Requiring additional infrastructure in other silos acting as more localized repositories

• Deploy application acceleration and WAN optimization services such as Cisco Wide Area Application Services (WAAS) Version 4.0

Solution

Cisco WAAS allows HCOs to consolidate distributed servers and associated infrastructure while leaving behind a component, a Cisco Wide Area Application Engine (WAE) Appliance, which can optimize TCP and application protocols to minimize WAN usage and improve performance. A remote user who performs an image query or retrieve receives LAN-like access to centralized server resources independent of actual location, and the requests use the network capacity in an efficient and high-performance manner. Wherever located, the user has consistent application performance and image transfers. The Cisco WAE Appliance provides local optimizations for file services protocols used by PACS and DICOM to reduce redundant transfer of unchanged objects. The WAE also provides WAN optimization capabilities to ensure that only byte-level changes traverse the WAN-doing so efficiently thanks to the TCP optimizations provided by Cisco WAAS.
Cisco WAAS addresses important operational metrics:

• Imaging report turnaround time

• Improved radiologist satisfaction

• Image integrity and delivery

• Centralized server and storage infrastructure

• Improved application responsiveness

• Minimization of WAN bandwidth consumption

Image delivery is accelerated through CIFS and NFS protocol acceleration and safe object caching. Building of the CIFS and NFS object cache is based on requested image segments, and metadata such as file attributes and directory information. Cisco WAAS eliminates the delays in transporting files across the network by intelligently serving data from the Cisco WAE when the files have not changed on the central file server. If a file changes, transmission of the updated file is first optimized by Cisco WAAS Data Redundancy Elimination (DRE), which is an application-independent cache of data from previously seen TCP byte streams. With DRE, Cisco WAAS can safely eliminate the need to transfer previously seen blocks of data, regardless of the application protocol from which the data was learned. The file protocol acceleration capabilities of Cisco WAAS minimize the workload placed on the network and the central archive, thus helping HCOs scale their existing centralized infrastructure and free network resources for other purposes or operations. Cisco WAAS helps ensure data integrity, which is critical for HCOs, by validating the state of the cached file each time the file is accessed, propagating critical messages (such as file locks, security, or write operations) to the central file server, and verifying message integrity.
Cisco WAAS further reduces WAN bandwidth demands-and improves imaging operations-with several important bandwidth optimization features. These features directly address reducing image transport requirements by caching common image data and minimizing DICOM protocol overhead traffic. They also improve bandwidth allocation by reducing the number of DICOM messages that are needed to complete an operation, reducing the amount of bandwidth consumed and WAN round trips necessary to transfer image data, thus dramatically improving performance for the remote user. Cisco WAAS maintains packet-network compliance through transparency to help ensure that other network features that are critical to HCOs, such as quality of service (QoS), NetFlow, access lists, and firewall policies, are fully preserved, thereby helping ensure investment protection, because network features implemented to control traffic and align network resources with business priority are preserved.
By enabling a centralized infrastructure while maintaining or improving remote-user performance, HCOs can realize more efficient use of storage and simplified data protection and disaster recovery implementation. The amount of unused storage and redundant capacity is minimized, thereby enabling a more cost-effective and efficient, centrally managed storage infrastructure.

Figure 2. Cisco WAAS 4.0 Optimizes PACS and Digital Imaging Networks

Cisco WAAS Performance Improvements for PACS

Cisco WAAS provides significant performance and bandwidth savings for PACS applications. Testing was done using a 3 Mbps link between a remote web client and a regional PACS device. The results shown in the graph below, figure 3, show that Cisco WAAS can provide up to 98% improvement in response time. Figure 3 shows performance improvements of Cisco WAAS when accessing PACS CT scan over a WAN link with 3 Mbps bandwidth. Cisco WAAS Data Redundancy Elimination combined with LZ compression removes redundant data from the network while Cisco TFO optimizes the TCP transfer of non-redundant data. Further, Cisco WAAS' Layer 7 CIFS Acceleration mitigates application layer latency and saves bandwidth for redundant transfers of the same objects when using Microsoft CIFS/SMB file servers. The combination of these technologies dramatically improves the performance of PACS enabled applications.

Figure 3. Cisco WAAS 4.0 Performance Testing Using a 450-slice CT View

Summary

The DICOM standard was created by the National Electrical Manufacturers Association (NEMA) to aid the distribution and viewing of medical images, such as CT scans, MRIs, and ultrasound. It eliminates x-ray film and embraces the latest digital imaging technology to quickly reduce overhead costs in labor, film, and chemicals. Cisco WAAS helps healthcare providers consolidate these large images, improve throughput, and reduce WAN congestion as data is being transferred across geographically dispersed sites. With Cisco WAAS, access to centralized PACS, radiology information systems (RIS), and other systems is accelerated, thereby improving productivity and patient satisfaction.