NetForum uses cookies to ensure that we give you the best experience on our website. If you continue to use the site, we'll assume that you are happy to receive these cookies on the NetForum website. Read about our cookies.
NetForum Community
Learn. Share. Optimize.
Log in | Sign up now | Submit content | Contact
Go to similar content

High speed, leading position

White Paper
Philips CT Marketing Philips Healthcare

Introduction

"Our Extended Brilliance Workspace shows an almost perfect linear improvement in performance when running on the latest Quad-Core Intel® Xeon® processor 5400 series." - Baruch Sabbah, Workstation Program Manager, CT Engineering, Philips Healthcare

Company: Part of Royal Philips Electronics, Philips Healthcare develops best-in-class imaging technologies for the healthcare sector to enable faster and more accurate diagnosis and treatment.

 

Product evaluated: Quad-Core Intel® Xeon® processor 5400 series.

 

Challenge: To improve the simultaneous visualisation and pre-processing capability of the Extended Brilliance Workspace (EBW), enabling it to process even larger amounts of raw data in a shorter timeframe

 

Results: The EBW shows a 100% increase in performance running on the Quad-Core Intel Xeon processor 5400 series with 45nm Hi-k technology compared to the Dual-Core Intel® Xeon® processor 5100 series. Multi-core technology means the EBW can perform heavy computation pre-processing in the background, whilst the user views another patient's data with almost no performance degradation.

 

Impact: Improved workstation throughput means the EBW can produce higher-quality 3-D images in a shorter timeframe, enabling physicians to make more accurate and speedy diagnoses. The ability to produce even better computer tomography systems enables Philips to maintain its leading market position.

 

Next steps: Philips plans to migrate its EBW to the Quad-Core Intel Xeon processor 5400 series platform and continues to work closely with Intel, aligning its product development to the Intel® roadmap.

 

Challenge

Philips Healthcare is a leading provider of advanced computed tomography (CT) scanners to hospitals around the world. Its range of multi-slice CT systems, known as Brilliance* CT, enables clinicians to carry out routine clinical exams, as well as advanced cardiology, oncology and CT/positron emission tomography (PET) imaging.

 

Brilliance CT solutions consist of complex medical hardware that scans the patient, producing up to 1,000 pictures, or slices, per second. Software running on either a server or workstation then processes this vast amount of data to generate detailed 3D images of the body's internal structure that can be used and interpreted by medical staff. Recent advancements in scanning technology have resulted in an explosion in raw data, generating large datasets that take time to process.

 

To maintain its leading market position, Philips is continually investigating ways to improve the performance of its Extended Brilliance Workspace (EBW) - the number one CT user environment for advanced visualisation. If the EBW workstation can process more raw data, it generate even higher quality images. What's more, if the workstation can process patient data in the background whilst the user simultaneously views another patient's data in the foreground, throughput will increase and physicians will be able to make faster, as well as more informed, diagnoses.

 

Deployment

The Extended Brilliance Workspace (EBW) is currently powered by the Dual-Core Intel® Xeon® processor 5100 series. In an effort to further improve the workstation's image rendering capability, Philips was keen to investigate the performance benefits of switching to Intel® quad-core technology.

 

The development sites for Philips Healthcare and Intel are both located in Haifa, Israel, which has enabled a vast interchange of technological knowledge. For many years the companies have worked closely together to align Philips' software development to the Intel® roadmap to reap maximum benefit from Intel's technical advances.

 

The current EBW software is multi-threaded, meaning that tasks can be executed in parallel on individual processor cores. In theory, multi-threaded software should show significant performance improvements when running on four rather than two cores.

 

Working in cooperation with Intel application engineers, Philips set out to prove this hypothesis.

 

Results

Philips tested its Extended Brilliance Workspace (EBW) software running on the Quad-Core Intel® Xeon® processor 5300 series and the second-generation Quad-Core Intel® Xeon® processor 5400 series. The EBW performed 18% faster on the Quad-Core Intel Xeon processor 5400 series with 45nm Hi-k technology than on the previous-generation Quad- Core Intel Xeon processor 5300 series with 65nm quad-core technology.

 

Intel application engineers then translated the Philips software code to the streaming SIMD extensions (SSE4) instruction set, optimising it to run on the Quad-Core Intel Xeon processor 5400 series. After this modification, performance increased a further 10-12% - showing an overall improvement of 20-30% against the Quad-Core Intel Xeon processor 5300 series.

 

Compared to the Dual-Core Intel Xeon processor 5100 series, the second-generation Quad-Core Intel Xeon processor 5400 series with 45nm Hi-k technology showed an impressive 100% performance increase - proving Philips' multi-threaded software runs twice as fast on four cores than it does on two cores.

 

Multi-core technology also enabled the EBW to perform heavy computation pre-processing in the background, whilst the user simultaneously viewed another patient's data in the foreground - with almost no performance degradation.

 

Impact

The two-fold performance increase means that the EBW can process data more quickly. Doctors and clinicians can make more rapid and informed diagnoses, improving the overall level of patient care in key areas like cardiology and oncology.

 

In addition to processing algorithms for image rendering, the EBW also processes large algorithms for computer diagnoses. An increase in cores from two to four means that these too can be processed in parallel, further speeding up the diagnosis process.

 

Another major benefit to Philips of the second-generation Quad-Core Intel® Xeon® processor 5400 series is that it allows the image rendering to take place within the CPU, eliminating the need for a graphics card. This means that the performance of the EBW is no longer restricted by the memory available within a graphics card, and is no longer slowed down by the transfer of data between the CPU and a graphics card. It also means that it is much simpler for Philips to make ongoing enhancements to its software that can now run on multiple platforms rather than just those compatible with certain graphics cards.

 

Together, all of these improvements mean that Philips can offer physicians all over the world leading-edge computed tomography solutions, helping it to retain its leading position in a highly-competitive market.

 

Future

As a result of the evaluation, Philips has made the decision to migrate its Extended Brilliance Workspace (EBW) workstation to the Quad-Core Intel Xeon processor 5400 series with 45nm Hi-k technology.

 

It continues to work closely with Intel, aligning the development of its Brilliance* CT systems to the Intel roadmap. Together they are currently evaluating the benefits of migrating Philips' Brilliance Everywhere - the client server model of its CT user environment for visualisation - to the latest Intel® multi-core technology.



This content has been made possible by NetForum Community.
Share this on: Share your link in twitter Share your link in facebook Share your link on LinkedIn Print Rate this article: Log in to vote

 
Rating:
Votes:
0
Views:
1360
Added:
Feb 20, 2009

Rate this:
Log in to vote
 

White Paper
Brilliance 16-slice, Brilliance 40-channel, Brilliance 64-channel, Brilliance 6-slice / 10-slice
3D, abdomen, Body, Cardiology, chest, Extended Brilliance Workspace Version 3.0, Extended Brilliance Workspace version 3.5, Head, image quality, Neuro, Oncology, Pelvis, throughput, Vascular
 

Clinical News
Best Practices
Case Studies
Publications and Abstracts
White Papers
Web seminars and Presentations
ExamCards
Protocols
Application Tips and FAQ
Training
Try an Application
Business News
Case Studies
White Papers
Web Seminars and Presentations
Utilization Services
Contributing Professionals
Contributing Institutions
Become a Contributor